I stopped using Cursor when I started getting comfortable with Codex/Claude. Cursor is just annoying with the constant popups and it's just not as good. Now my workflow is to use my normal editor, add a todo describing what I want, and then ask Codex+gpt-5.5 to implement it. It absolutely nails it. Using codex is so much more like working with a partner vs the noise and annoyance of Cursor.
That said, I think we're in a narrow window of time right now where any of this matters. Prompt "engineering" and working around your tools will be over in a year or so.
Fwiw I am a c/c++ systems engineer. I think anyone mentioning anecdotal experience like this should clarify. Maybe frontend JavaScript folks have a totally different take and that's expected.
I use cursor 8+ hours/day at work, and have full (and effectively unlimited) access to Claude Code and Codex - tools which I also use personally. I suspect that your "constant popups" were when you were using the editor - a mode that I'll confess I haven't touched in 3+ months.
Workflow in Cursor is actually awesome - I'm a little outdated in how I use it - I still establish goals/objectives, rather than managing the loop which does so - but if you can think broadly enough - I find it's pretty efficient.
Key things I like about Cursor (and I recognize I'm dating myself a bit here):
- Plan Mode is really solid - I shift-tab, have it go create the plan using whatever insanely expensive SOTA model is available - I will usually spend 5-10 minutes on the Plan - review it, maybe even tweak it a little. (though 90% of the time it's fine out of the gate)
- Ability to select any model for every task - I'll switch between Opus 4.8 High/xHigh/... I'll even switch to 1M context for the planning phase upfront.
- It does an *excellent* job managing permissions and looping the agents and spinning up sub-agents for you - you set the goal, run the plan mode - and then let it churn for however long is required - pretty common to have a 30-45 minute run and come back to a fully created/tested product.
The nice thing about Cursor (and honestly Claude Code, Codex) - there isn't really any "prompt engineering" involved. You just say, "Go Build me x - it should have y,z features - and build it in golang for me" - and that's it - the 3-4 page Plan comes back - usually pretty credible - and then you click "build.".
> there isn't really any "prompt engineering" involved
You should make an experiment; take someone who never used any LLMs or agents, and tell them to use it for the first time in front of you, and tell them to build something like a calculator program or whatnot. Bonus points if they're ICs or at least not-managers.
I think there is a lot us engineers take for granted, when it comes to communicating via text, how to state things clearly and what we think/reason when we read things. A lot of people don't have those "skills" innate, and the first time they use LLMs, they basically don't know how to interact with them, until they realize what they're able to do and not. Then they also learn what to say to steer the model into the right way, this is quite literally a "prompt engineering" skill they're now learning.
You don't even have to go outside engineers. I have teammates that get very little out of Claude Code because the way they integrate their own knowledge doesn't allow them to think of what Claude might not know. They'd say a task was impossible with the tooling, and I'd get instant answers, because I understand what is weird internal business logic sitting 6 repos away, and what is knowledge claude has by default. I can commit Claude.md files for them, but I have to include EVERYTHING, because otherwise they'll let Claude make assumptions and waste minutes, if not hours.
It's a big part of what, in my experience, is separating the very good engineer from the iffy one: Do you have a good mental model, and can you put yourself in the shoes of people sitting in a different mental model? It makes you a better dev, and even more so when it comes to AI tools, which have their own kind of alien brain.
Coding LLMs are distilling developers. It's like the old experiment where you have someone write down the steps to make pancakes and they don't tell you to crack the eggs before adding them to the batter: it takes a particular mindset to be able to make a model of what is supposed to happen and deconstruct that to the level appropriate for implementation.
Until now, the actual act of writing code: terminology, syntax, etc. was a significant hurdle, and that underlying mindset was a very useful, but missing in a surprisingly large number of developers, skill.
Now with LLMs doing the work of "translate this into code," increasingly the only thing that matters is that exact ability. And developers that don't have it or can't develop it won't be developers for long.
Thanks for putting into words what I have been seeing a lot at work and haven't been able to put my finger on. We tend to have quite diverse _workflows_ between devs at my company, and success seems to correlate with injecting better context earlier in the process.
I like to chat with Claude about how to approach a given problem, bring in extra context, etc, before even really drafting up a plan, while other people dive into implementation immediately and go on wild goose chases.
90% of the time we end up in the same place in roughly the same amount of time, and there are obviously tradeoffs to spending more time planning vs implementing. I'm oversimplifying as well.
I couldn't agree more. Socratic methodologu, domain modelling, systems thinking, pipes-and-arrows problem solving etc. These are the skills that get real work done in coding agents these days.
I'm sure that explains some of it but I really don't think it explains most of the people who have been AI-pilled in the last nine months. There was no amount of context I could give GPT-4o that would make it a net benefit to use that for agentic development. I tried it with quite sophisticated prompt systems and much simpler ones, compendiums of code & business analysis and sparser ones. Yet it just wasted my time - still there were people using Cursor with that model and saying it was life changing. I didn't have that experience until Opus 4.5 - its possible I could have had it earlier but that was when I happened to try it again.
I think many of the people who have become "AI Pilled" (I'll include myself here) had it happen in the last 3 months. Even over the Christmas break, when the Wiggums loop got so much coverage - I still wasn't that blown away going into January/February- 50%+ of the time I'd just write the code myself. I like coding.
But - I don't know if it was April, or May - but very recently - the coding harnesses paired with decent SOTA models like Opus 4.8/GPT 5.5 - just started showing a lot more consistency, and completeness, and sometimes downright clever behavior - that they started to become way more useful.
Just one out of hundred+ examples - I gave Claude Code (Opus 4.8 High) a complex task that involved consul, vault - but I had neglected to give it sandbox permission to download from hashicorp.com. So - it created a entire test harness that simulated both the behavior of Vault and Consul - created all it's test cases, verified that they passed - and when I came back 40 minutes later said that it was all done.
It's test harnesses so accurately simulated the behavior of Vault/Consul - that on first try - no refactoring whatsoever - all of the protobuf/AESGCM/API behavior (that has varied significantly between versions) - worked.
This was something that would have taken me, someone super super familiar with the code and tools and APIs - a minimum of 3 solid days of work - and that would likely involve hundreds of attempts and refactors as I unwound all the weird encryption and packaging layers. It zero-shotted a full solution without having an API to test against
If these agents actually have an actual test-harness - It's honestly hard to imagine what they can't do - subject only to imagination and budget at this point.
Speaking personally - something changed Between January and, Let's say May - in which instead of seeing these things as mostly interesting technology demonstration, in which the flaws outweighed the benefits - I now genuinely think they are the future of programming. I'm dubious that I'll write much software manually in the future - beyond what I do for personal pleasure.
Asked to write a driver for macOS for some thing that didn't have macOS support, GPT-55 found Linux OS firmware on the vendors site,
downloaded it, ran binwalk, extracted out the driver, got halfway to reimplementing it on macOS with barely any help from me. I did need to dive into it somewhat to get it across the line, but it showed some ingenuity along the way.
Some people "got" LLMs back in 2022, others needed it to evolve a bit.
It's not unlike computers. I started using them back in the 90s and absolutely nobody I knew was interested, while today everyone carries one in their pockets...
By that same logic (and I’m agreeing with you as of now), engineers shouldn’t get too comfortable treating “being good at text communication” as a lasting edge. With how quickly agentic coding is evolving, it’s worth considering the possibility that many of the prompting and steering skills we view as valuable today could become far less important in a matter of weeks or months.
Recently I have the SEO guy governing the mostly static, public site with Claude Code. He loves it but you would never imagine the level of mental illness Claude comes up with. If it were an employee I’d literally throw him out the front door, labor laws be damned. And as always, every insane thing it does is some direct echo of its concept and training.
But what's the $60B differentiator here? There are so many similar tools out there. I generally use Opencode, but also Claude code, antigravity and sometimes Kilo code on VS Studio. How can cursor be worth even 10% of 60B?
I don't know what cursors market share is but it feels like 20-25% to me. That is not worth nothing. Then;
1) The data they have flowing through the system that enabled them to build composer (which is much better than stock kimi 2.5) and is presumably allowing the training of a new model on space Xs compute.
2) Cursors new 'github' replacement.
3) Enterprise sales/traction
If you look at all of these together, it's not implausible that they end up mostly 'owning' coding in 5 years time. If they replace GitHub with something more compatible with agentic coding and bring it into their whole ecosystem providing cloud and local agents, PR review and own frontier coding model.
It's specialised vs 'borg' isn't it. One way of thinking is that the world is owned by Anthropic/OpenAI and coding is just one of many things their model and software does. Another view is we have a 'coding with LLMs' company that specialises in this field of endeavour. Hard to say which wins, but I think they have a shot.
Personally my only objection to cursor is that it's more expensive. That's it, otherwise it is great to be able to choose say GPT-5.5 when I want to work on backend and Opus when I want to work on front end. Great to have PR review built in. If they were able to get composer 3 to as good as GPT5.5 / fable at the price of composer 2.5 they'd be winning on price again.
> If you look at all of these together, it's not implausible that they end up mostly 'owning' coding
They really need to change their trajectory then?
And regardless being owned by xAI, a failed AI company which turned into a datacentre operator probably won't help them to achieve that.
> Hard to say which wins, but I think they have a shot.
The market for "coding harnesses" and "AI IDEs" is already oversaturated and they are effectively a commodity at this point, you can use any of them with any provider more or less interchangeably.
> They really need to change their trajectory then?
They need to step up progress sure.
> And regardless being owned by xAI, a failed AI company which turned into a datacentre operator probably won't help them to achieve that.
I think near unlimited access to compute is exactly what they need to train a frontier level coding model and serve it cheaply and profitably.
> The market for "coding harnesses" and "AI IDEs" is already oversaturated
I think my entire point was that it's not just a AI IDE. It's a coding focused model (currently Composer 2.5, soon hopefully something better), a Github Replacement, PR review/Bug Bot, Cloud Agents and so on and so forth. It's a ecosystem. An enterprise signs a MSA with you and gets everything they need all in one place.
Yes because Grok failed and they now have "unlimited" compute they can sell to other. I mean you are right that if they did X, Y and Z they could be very successful but their is no indication that might happen. In any meaningfully way seems like Cursor has peaked a while ago.
> An enterprise
Well either they are the type of companies which just buys whatever Microsoft is selling OR they let their developers to mostly pick what they feel is the best tool for the job on their won. I don't think there is that much in between (and its a cutthroat market e.g. GitLab)
> a Github Replacement, PR review/Bug Bot, Cloud Agents
Those things are a dime a dozen, you can vibe code them in weeks/months and there plenty of options on the market already. Well not Github of course, but there are various reason for that which have little to do with product quality and features (not that I think there are many companies which could build a meaningful GH replacement in a realistic time period despite its many flaws).
I just don't really see a huge income stream for dev tools companies (just like there never was) they can skim of something from the top by reselling AI models (generally at zero or negative margins..) but that's not the most lucrative business model when you have no real moot.
By not succeeding? It's an also ran, a closed proprietary model which is behind Anthropic, OpenAI, Google and a a bunch of Chinese companies, how do you make money with a produce like that? (besides the absurd IPO of course...)
For a lot of people, Grok is the first AI they got to use through Twitter. Grok does get quite a lot of usage, and isn't out of the game - coding tools aren't the only use case for AI.
Google glass has been discontinued? Besides, many people use it on Twitter everyday. Usage is not limited to what you can see on the Openrouter dashboard.
these users are probably losing the company money.
the failure is in converting regular people into actual ai product consumers. Companies are realising that the money is not in regular consumers but in enterprise and they are not considering grok as a serious alternative.
if anything, the name, the branding and the x/twitter affiliation has hurt adoption from money makers rather than help it.
so yes, people know it, but no one is willing to pay for it
Depends, Grok stimulates engagement and pushes to stay on the plaform and feed it data. If anything, it helped justify a massive valuation for SpaceX, which is a metric of success for most corpos.
It helped the valuation but as just like SpaceX hallucinations about the space data centres. Doesn't mean its not a crappy low end model itself. Btw is Twitter even making any money?
There's a HN article and discussion about Anthropic expanding to use Colossus 2. https://news.ycombinator.com/item?id=48214017 I think it's fairly clear that grok isn't using as much compute as expected.
So far seems like none of those use cases have generated meaningful income streams? The consumer/non-developer market is mostly dominated by OpenAI and Google anyway...
> The market for "coding harnesses" and "AI IDEs" is already oversaturated and they are effectively a commodity at this point, you can use any of them with any provider more or less interchangeably.
Yes and no. I've used a few different harnesses with closed and open models and there is definitely something going on that makes some harnesses work better than others. Many of the differences are hard to pin down and some are things people don't care about. But I wouldn't say they are commodified just yet.
1. Memory use. I have colleagues complaining that Clause Code uses several GB of memory. Meanwhile I haven't heard about that regarding codex or goose, or even opencode for that matter.
2. Suitability for local models. When you use Anthropic models, you use Anthropic as a provider. They can have software between the model and your harness that will fix issues with the model. One notable thing that even the best open weights models struggle with is broken tool calls. There is a lot that a harness can do to fix broken tool calls when working with a straight up ollama running a raw GGUF file.
3. Ease of use with non mainstream models. OpenCode has GREAT coverage of models/providers. Goose, less so as it relies on people to set up their own anthropic or openai compatability settings. e.g. Zed doesn't let you use Z.ai (which, if you speak British English, sounds ironic because "zed ai" isn't directly supported by Zed the editor).
4. Worktree support. Opencode and probably all the TUI harnesses works in a local directory - so you need the terminal to be in the worktree. Zed, however, works centrally on your git repo and tracks the worktrees so you can bounce around your work in a single window.
Of these, '2' is maybe the most important one but also the hardest to pin down as a feature. '3' is a one time cost. Of course '1' could be a blocker for someone using a macbook air or neo.
I agree, Composer Fast 2.5 is getting really good. I started using it for a personal project after I had to switch from Sonnet because I hit the API limits, and I was surprised by how good it has become.
I believe they have some very good training data because of all the data generated by people using the service.
This is the same data they used to finetune Kimi K2.5 to make their newer Composer models, which benchmark substantially better than Kimi K2.5.
I've heard they also want to build their own base models, which will also benefit from their large amount of high-quality training data. Which will solve Grok's model quality problem.
This is all unsourced conjecture of course. But it's what I've heard.
Also from what I understand (not my day job) we're now at the point where the post-training tuning (RLHF etc.) is increasingly important since pre training no longer scales.
So it's not really fair to call it "fine tuning", it's an important part of building a coding model in 2026, and cursor have done a pretty good job with Composer
they are paying for marketshare/customer base. Cursor has a good chunk of it.
xAI overbuilt their data centers - they can't find paying customers for them, that's the reason they made deals with other companies like Google to use their own datacenters.
Cursor has the opposite problem of not having enough capacity. So this works well for them together.
Weather it's worth it - if you beleive that AI will solve every problem then having a piece of the pie early on might be worth it.
Remember how when google bought youtube for 1.65 billions people thought they are crazy? Or when facebook bought instagram.
60B is a crazy number but might be worth it for someone fighting for world dominance :)
xai is on the line to delivery capacity they already sold to Google and most analysts think they are 50/50 on actually meeting it.
the only proof they have capacity is that musk claims all the money they are burning is going to datacenters and gpu (mostly because if he put it on anything else the lie would be obvious)
Or are they paying for talent? It seems like xAI is sorely lacking in talent, most likely due to the CEO and folks' aversion to him. By throwing around some SpaceX monopoly money he can trap some talent with retention clauses and try to invigorate his failed AI business.
I think the argument for Cursor is that it's the dominant tool that enterprises are using for coding, so the theory is Cursor wins that as the "model agnostic", it has a phenomenal Enterprise Sales Team.
From a valuation model - $4B ARR with rapid growth, and the ability to shift traffic to internal models (honestly, massive amount of the time "composer" - their internal model is fine, and obviously going to get better). Say 17x Multiple which isn't unheard for a rapidly growing Startup with solid future structural profit elements (moving to internal model) - that gets you to $68B.
> so the theory is Cursor wins that as the "model agnostic"
But there are many model agnostic harnesses out there: OpenCode, Roo, Cline, and many others. And even Claude Code can be setup to use non-Anthropic models.
As a Cursor user, I don't have to have thought about the providers behind the compute - I get name brand Claude, or cheap Kimi, or Grok, and it's all got roughly the same agentic experience, and only one bill. Enterprises love this.
If you resell something worth $5 for $5 while having to pay for R&D and operating expenses that's not exactly comparable with a company that's selling actual products.
> Say 17x Multiple
On an extremely low margin business it is, yet again that wouldn't be the stupidest thing in today's market.
It can't as long as there is plenty of AI without it.
The real differentiatior is that if $60B today turns out to be all thrown away in a worst-case scenario, it would be easily more affordable and there would be less negative impact than $47B at the time if it was all thrown away on Twitter.
We’re in the new era where startups boast about and bought based on revenue and not on just a number of users with unclear path to monetizing as it had been for the previous couple decades.
We can also note that we see Thrive Capital (Kushner) again in a win.
Where else are you going to get access to a real-time fresh high quality stream of human intelligence to grow your baby AGI? You can’t buy Codex, Claude, Copilot, so what’s left?
How are you switching between like 5 different editors lol. Bro sloppers will do anything to get their fix. Like the old people at the casino switching slot machines all day based on some occulted understanding that only they think they have.
There is most certainly still prompt engineering involved. How there can be both the responsivity to different cues like "plan this", "write this", "analyze this", "defend this", "poke holes in this", but not responsivity to the various terminology you provide in your explanations of "this", where to get information about specs/standards/requirements, what details I care about, and therefore can't compromise on, vs what details I'm willing to accept whatever the top reddit post from 4 years ago recommends.
I don't see how these systems can have the ability to be effectively expressive about all of the minutia, and not have all of the various different possible expressions lead to vastly different outcomes.
I think all of the cues that you just described are in the plan.
For example - I might (real world example from this morning):
"Create a script that installs hashicorp vault and consul, store the data on consul. Then create ahelper script that will fill the vault server with sample data. Add HTTPS support. Now write a framework that reads and decrypts the encrypted data in consul. Support old (pre 1.3) and new (post 1.3 vault). "
That generates a 6 page plan using Opus 4.8 w/1mm context, including notes on what to prioritize, what format to create the scripts in, etc... (My cursor guidance already has a couple months of hints as to what I want in terms of scaffolding unit tests, canonical linux, performance, security, etc...)
That 6 page plan is the "Prompt" - but it's entirely generated by Cursor/Opus. It's there to tweak if you want to emphasize, or provide some taste - but, honestly - it probably does a better job than I would - so ~90% of the time I just accept the plan as is.
I would say prompt engineering, in the sense of people claiming you need to include in every prompt magic incantations like "You are a senior engineer from a superintelligent alien species" and "take a deep breath and make no mistakes" doesn’t really do that much for everyday work I feel or they are all already included in the system prompt maybe. I reckon it can still edge out a few percentage points in automation.
What actually matters is the ability to communicate well in general, not anything LLM-specific. Being able to state what you want clearly and unambiguously, and having a sense for what additional information you need to dump, even when the other side claims they already have everything they need.
> You just say, "Go Build me x - it should have y,z features - and build it in golang for me" - and that's it - the 3-4 page Plan comes back - usually pretty credible - and then you click "build.".
What you're describing seems like a workflow for building toys only. There's currently no reality in which someone would actually know what the y,z features are before making them. A plan generated in 5min would likely suggest a suboptimal solution compared to what a good solution would look like (which might take a year or two to figure out, for a human, so still a week or so for SOTA models if at all possible). Building something in golang is cute, but hard to be convinced until more novel applications are being generated from prompts.
The data submitted by Cursor's users tho, that seems to be very valuable.
Yes, I tried to use Cursor as an editor. Terrible idea in hindsight.
So your workflow now looks like mine except I prefer a different editor and only use the latest and greatest model so Cursor basically offers nothing over Codex.
I disagree about prompt engineering, but it's one of those things that probably varies because of what language you use, what problems you solve, and the degree to which you care about the output. Unless I'm writing tests, I keep AI on a very short leash because I'm writing critical code used by a very large number of users. I have noticed big differences in output quality depending on how I steer AI. Without steering, it will happily leave in dead code, change the use of variables so they need to be renamed, assume or fail to assume invariants, etc. As I said in another comment, I think we won't need to do that for very much longer, but right now it seems essential.
You nailed it - in fact, most of Anthropic's early revenue came from Cursor - much of claude code programming components is essentially a feature copy of Cursor, so it makes sense they are similar.
Cursor does have it's own model - it's a heavily reworked version of KimiK2, called "composer" - that I use a lot of the time when I have fairly straightforward tasks that don't require a lot of exploration or independent thought. Lot cheaper - the Input/CacheWrite/CacheRead/Output costs of Opus 4.8 are $5/$6.25/$0.5/$25 per mm tokens, vs $0.5/-/$0.2/$2.5.
Not trying to be funny but seriously, if these tools can produce a tested 'product' in 45m, shouldn't we be seeing millions of them out there? I mean how far are we from a fully AI built Oracle ERP or even a notepad or helix?
- Note - that my "product" was about 3,000 lines of code - so tiny. But https://metr.org/ should give you some insight into the complexity the models are capable of.
- you have to be able to imagine the product. If I have the time, and energy, to imagine what I want - the model will build it. Here is an example of a much better programmer than I and something he wanted built - https://www.boatbomber.com/blog/claude-fable-5
- These are the first drafts. On average - any complex system needs about 10 years and at least 1000 active and enthusiastic about reporting users to really get robust code. Writing if via LLM doesn't (at least so far in my experience) help that much in reducing bugs if you were previously following any semblance of TDD. Lots of bugs in the code - the products you listed above have literally tens of millions of years of user experiences and bug reports that got them to where they are today. No silver bullet yet - just faster, less effort - and it enables non-technical people to create (still buggy) products.
Millions of produced verified software engineered products in 45 minutes in the likeness of Oracle ERP or notepad++, helix are small potatoes when you see the unbounded ambitions of SpaceX in full.
The end point may squeeze quality of operations at the subminute time span for ground control environment seriously launching Starship rockets one an hour, for example.
You absolutely don't. I use all three products. My preference is Claude Code for my personal project. The one at work is kind of sandboxed off - but does have the benefit of an MCP for every enterprise service we have (Kibana, Victoria Metrics, Grafana, Jira, etc...) - which is nice.
Over time - I expect Composer will be cheaper than Opus 4.8 - but the nice thing about Cursor - you can flick between models.
And (this is purely a personal thing) - I really like the extensive collection of "Plans" that cursor tracks - there isn't really a similar thing in Claude Code - but I really like the Claude.AI interface for everything else. It's also a much better general knowledge agent - the Cursor Chat interface isn't as nice.
I’m not sure what you’re on about. I had Claude doing swarm engineering using different models. It would write specs that haiku would implement, it would check itself etc etc. with a simple phrase it goes into planning, multi agent mode, and chews on a problem until it’s done. It’s pretty autonomous.
Maybe you haven’t looked deeper into what modern Claude can do?
The Different Model approach is where from tasks to task - I can switch from Opus 4.8, GPT 5.5 and (very often) composer 2 at 1/10th the cost.
It's not perfect, btw - to some degree you are at the mercy of which models they support - currently only 27 from Gemini, OpenAI, Anthropic, Grok, and Kimi (Just K2.5) - presumably because they have commercial arrangements with them. The "Bring your own Model" model requires you plunk in your API key - which sucks. And only one at a time.
To the best of my Knowledge, Claude Code only supports one model at a time if it's not one from Anthropic (which will use the the entire suite of Anthropic Models depending on the task) - and you have to override it to a single model with an environment variable at startup - no ability to flick between models from task to task.
Depending on your workflow - you can save 70-90% on costs just by chosing a reasonable model for really extensive tasks that don't require thinking, max context, etc....
Different models aren’t subagents - they’re completely orthogonal. I use Gemini subagents for code review in cursor, but mostly use gpt for actual coding.
When I first used Cursor, I hadn't used any of the "Vibe Code" tools out there, so it was pretty neat to have an assistant directly tied to the editor.
Once I learned how to use Codex, I just used a tmux split with NeoVim and have the effect I wanted. I haven't felt compelled to use Cursor at work since.
I also work with C++, and I use Codex (desktop) which writes 99.99% of my code, plus Visual Studio, which is nice for reading and navigating code. For webdev I do VSCode + Codex.
I started with Cursor back in the day, but switched to Claude Code and then Codex when Cursor got too expensive.
If price wasn't an issue, maybe I'd prefer Cursor only because I can easily switch between models. But that's it. I always disliked the "accept/reject" workflow in cursor, but that's probably optional nowadays I guess?
I love the accept reject flow because I still constantly have to stop AI models from writing awful architecture or reimplementing code we already wrote elsewhere
Yeah, I have found the same. A lot of times it does get things right, but if it deviates man it can just drift hard.
For example, sometimes Claude just obsessively reads files and goes on massive tangents. Then when I stop it and ask, "why are you doing that?", it kindly apologizes and admits it shouldn't have gone on a tangent.
The token burn if I don't stop it would be quite high.
Granted, this might be because I'm not giving it optimal prompt/negative-prompt instructions though.
How is it different from Keep / Discard in other tools? I've been slowly converting my git repositories to jj locally because that gives me more granular fallback and mix and match options.
Well I tried CLaude Code for the first time in a while (I am building my own coding app www.propelcode.app so I can code on my phone when I take my kids to classes and such) and it literally ignored my question and suggestion and just kept coding away.
I hate the accept reject flow, because I want a conventional code review workflow where I can write comments on specific lines of code and maybe edit the code myself.
If I reject, then the AI will struggle to modify just the parts I disagree with, if I accept, the AI will tend towards adding code rather than updating the bad code.
At that point copy paste without agentic coding tends to work much better.
Fable makes any IDE AI integration almost entirely unnecessary. Claude one shots pretty much everything, and fixing any small errors is easier when just talking to Claude again.
Anthropic is going to offer better pricing using their agentic harness. Why pay more for less?
An IDE at this point is best as a tool for code review. They need to start building better code review tools.
I can't quite understand the "fixing small errors is easier when just talking to Claude" flow.
I tried having it write some tests today. It got very close to what I want, but picked a stupid set of input values (two fields that look independent that should only be used with related values). I thought about "how do I explain this" and then just went in and fixed it myself.
How is it easier to write "Okay, go back to testBlah and change xxx to yyy" versus clicking on XXX in the IDE and typing YYY by hand? Maybe if you had 500 faulty tests and were forbidden from using search-and-replace for some reason.
It makes sense when code generation is the limiting factor, but I end up with a lot of changes where the actual code delta is smaller than the necessary prompt to convince the bot to produce it.
Try the superpowers plugin, let it write a spec (what do you want?) and a plan (how is it implemented). Then let it implement the plan.
Review each step as much as you care. These things take time so you can just do other stuff while it’s cooking.
With proper isolation of projects you can easily have multiple sessions in parallel. I frequently have 4 to 8 parallel Claude Code sessions, each with whole trees of agents reproducing, speccing, planning, implementing and reviewing things.
For common mistakes, you can make it remember things or rely on reviews.
the reason to use cursor nowadays isn't the IDE (though it's helpful perhaps once a week), but how it makes running models from multiple providers trivial out of the box. I don't have to juggle keys or drop to a shell tool call, it supports calling out to e.g. gemini in a subagent natively. I have multiple models cross-reviewing plans and diffs as a matter of course.
claude code was seriously annoying with the flickering, maybe it's fixed now, I don't know.
cursor also has a (bad) cli if you need it, it seems it's mostly used to setup remote agents, but it does the job in a pinch.
OpenCode is miserable from a security perspective. Well clarification the plans they offer where they bundled in free models that train on your use. You are then left to use an OpenRouter which I find pretty flaky for at least the leading Chinese models.
I doubt most people use OpenCode coding plans, nor do they use OpenRouter. I use subscription plans from ChatGPT, z.ai, MiniMax & Xiaomi with OpenCode. It handles authentication with all of them seamlessly. I switch between models based on task/subtask and based on usage limits. You can get the most value out of a lot of these plans at their second-tier and they are often switching in value relative to each other, so it makes sense to arbitrage them like this.
Most of that switching is automated (oh-my-openagent - defaults sub-tasks to different roles, so for example I use MiniMax for explorer tasks and GPT 5.5 for deep design & review tasks, and GLM 5.2 for general orchestrator & most coding). If I hit usage limits it switches to a backup for that task. I'm not sure Cursor authenticates with all the subscription coding plans from all those companies - but if it does it can't be doing it any better.
“Well clarification the plans they offer where they bundled in free models that train on your use.”
Just what I said. They offer paid plans through their tool. Said paid plans are kind of a dark pattern where it’s not immediately obvious the models are training on your data. The harness is fine but that kind of business turns me off and I am usually pretty neutral about those sorts of things.
I still saw a lot of flickering in VS Code (I simply use CC as a terminal in VS Code, without the plugin) as of 2 weeks ago. I think it's a combination of CC bugs + Electron(?) rendering the VS Code uses for terminal.
Moved on to Zed (native Rust rendering) 2 weeks ago -> nothing flickers.
Sadly, with Fable 5 cutoff, I am actively exploring CC alternatives. Pi/OMP.sh works great as an agent (definitely better than CC). GPT is seemingly not as good as Opus, but with better agent and better skills, it probaly won't matter anyway. GPT lets you use any agent on Pro subscription.
Maybe flickering, but it's still broken in various ways. Only a few days ago I had an issue where the text I was typing was outside of the textbox frame. Resizing the terminal still maintained the broken view.
Also C++ engineer, but from my perspective, for large tasks, agentic coding is still lacking no matter how well I describe desired output. So in that cases I fall back to manual coding and cursor tab helps a lot with boring parts
I actually don't let AI take on large tasks beyond test writing and refactoring helper scripts/utils. I keep it on a very short leash for driver/middleware code since the quality bar needs to be extremely high for our codebase. Up until recently I didn't even trust it for that, but some experiments show it's fairly good and even detected issues outside of the refactored functions which I did let it touch. This is with a good amount of 'thought engineering' though where I try to think hard about how to emphasize certain factors and define the problem as best I can.
I've used AI but not through Cursor to implement a high-performance serialization library using expression templates. Non-trivial and few kLoC targeting a soft-realtime and critical system. Would you consider that a large task?
On the flipside, I enjoy Cursor now and came back to it after leaving it over a year ago. The 2.5 model is fast as hell and very good. And whatever harness they have it's terrific, great results. I also really enjoy the fact that I can open my website in the Cursor in-app browser and just click and reference stuff. It's a really cracked workflow. The models can only get better for them.
I would also add that Cursor's "Debug" harness is incredible. Hit "Tab" in the AI editor to Tab through the options (Plan, Multitask, Ask, etc.)
If you do any kind of on-device work, it will spin up a local HTTP log server, and pipe logs from your real device (phone, hardware, etc.) to the server and do realtime debugging.
Claude will mostly guess, have you copy + paste logs, etc.
> I can open my website in the Cursor in-app browser and just click and reference stuff.
I’ve never used cursor and have only seen it in a couple work lunch and learn demos. I’ve never seen that feature. I have a lot of use cases where I’m asking cc to move a widget down a little bit or make a data table full width etc. Being able to reference the actual UI would be useful.
Dunno, Cursor's agents are now more-less equal to Claude Code, just the workflow is slightly different. I like the IDE integration for some projects, allowing me to quickly inspect/review/change/search code, while running Claude Code/Codex/OpenCode/Pi/Hermes on different projects often with local models and it's mostly a question about your personal development style instead of inherent tool capabilities.
It is possible to use Cursor via ACP, so you can use it in any editor that supports ACP (notably the JetBrains IDEs). Our company went all in with Cursor and at the same time centrally disabled the AI functionality of JetBrains IDEs, but a pretty large group of developers (me included) were so vocal about wanting to continue to use our "old" IDEs that IT eventually relented and enabled the plugins needed to support Cursor.
You know you can open the same project in cursor so agent does its own stuff and then opens JetBrains IDE to do your code navigation etc. ?
I am pissed off by people calling Cursor an IDE … Cursor is text editor with AI agents bolted on. I still like what agents do and how the context is managed in Cursor but it is far far away from proper IDE.
Right now I think there is an edge to how you construct prompts and config files. There is a large difference between "modify f() to do..." and "modify f() to do... Review the current variables and make sure they are still used consistent with their naming. Look for unreachable and dead code. Examine callers and called functions for side effects from the introduced changes...".
I don't think that will make much difference in a year.
I'm increasingly convinced of the opposite. IMO Fable was pretty similarly capable for my day to day work as Opus.
I think there's a pretty good chance that we've reached the point of diminishing returns, for our specific use case.
There are still like a billion other (more difficult) use cases to be tackled, but I think "generating code" has gotten really good to the point where the other bottlenecks will prevent further exponential progress on this specific task.
It's already going away for me in a sense as I build up a library of AGENTS.md and Codex skills. I see no reason such things won't get baked in at the agent layer so that domain specific rules and such are automatically applied when appopriate.
I'm not sure if you're ahead of me or behind me on this curve, but fwiw, my experience has been that we have now encoded everything that is useful in the various markdown files and have reached the point of diminishing returns on this, with more powerful new models making noticeable but not revolutionary improvements as they come out.
You're essentially making the case here that your work is now automated into a set of one-shot actions that can be performed by an AI model and your job has become to selectively apply these actions. That says either a) we don't do the same work, and instead you're doing some kind of low level devops function that I've only ever seen in rare cases where a human isn't needed anymore, or b) you've vastly oversimplified the software engineering you're doing.
Sophisticated chain of reasoning LLMs like ChatGPT have baked in some natural language operations and they make it so i can create at a higher level of the language expression stack. But I'm still formulating my own expression. There's no conceivable path I can see where an improved model is going to be able to do what I do. I think that is clear from my ChatGPT threads at least.
I think you're reading quite a bit into my comment.. I'll try to respond to a more accurate response to my comment but I'm not going to waste time with this sort of response.
Yeah, uh, why would it go away? In what world do you completely surrender your ability to control the work product, the methods for achieving said work product, etc. That is the dream of a PHB.
Not OP, but I generally agree. Models are powerful enough now to reliably instruct other models. They don’t need fancy tools or IDEs, just the command line.
With deterministic workflows, type-safe languages and test suites, agentic loops pretty much “can’t fail”. They will continue until the types resolve, the tests pass, and the project requirements are deterministically met.
By that point it’s literally just a case of typing a prompt in to a text field, and waiting.
Same path as you. Went from $60 cursor plan (often exceeding it which costed more in API) to a limitless $100 codex plan where I basically say "read the markdown and implement the instructions". Deepseek also works quite well, surprisingly!
(FWIW Im mostly using python for OCR, LLM calls, data analysis..)
I think I'm late to the party with cursor but I don't use it as an editor at all, I keep VS Code open on another screen for that. All I do in there is agent sessions. I would be open to something else but all the comparisons I see are out of date and talk about the IDE a lot.
The comparison is with Claude Code and codex (and open harnesses like opencode and Pi). IMO they are both better, if you aren't interested in the IDE functionality.
The code suggestions. It's highly distracting and pulls me out of my flow. I know how to code and I don't mind typing. I don't need AI making trivial suggestions. I want it to do exactly what I tell it to do.
I like cursor, but I'm assuming they're talking about how it hijacks your tab key. It's amazing when it works, and infuriating when I just want to insert a damn tab!
Composer is fairly decent. Many people aren’t in the market for an IDE — and a subpar one at that —, but they could sell API access to Composer itself.
I think Fable gave a bit of a sneak peek into the future.
My objective KPI: for the few days I was using Fable (18hr a day), it would frequently push back against my design ideas and propose alternatives -- and they almost always felt better to me. Back to Opus now, still 18hr days - and I dont think it disagreed with me meaningfully even once since Saturdy. I consider myself and old hand -- and i think Fable really didn't need me to be very specific in my prompts, it would have done a good job regardless, or even despite my prompting.
Of course whether this is the future is anyone's guess. Maybe we will experience a butlerian jihad and there won't be any prompting whatsoever for completely different reasons :-)
The models are getting better at agentic coding, so over time using complicated harnesses and precise prompt engineering to attempt to squeeze out an extra X% performance will become irrelevant as the models approach expert-level performance. The bitter lesson in miniature.
>> Fwiw I am a c/c++ systems engineer. I think anyone mentioning anecdotal experience like this should clarify. Maybe frontend JavaScript folks have a totally different take and that's expected.
Similar experience, having transitioned from full-stack to a dedicated C/C++ stack, learned/experienced firsthand that there is no one-size-fits-all tool.
multiple organizations I contract with have killed their cursor enterprise plans over the past several weeks
to me, this seems like the perfect time for Cursor to exit and even "Q3 completion" is too late. Deal just needs to close. Fortunately Q3 completion could mean July 1st too
I taught myself assembly language from a book on a 286, I cracked games with SoftICE as a teenager, tried out every Linux distribution in the 90s, and have been developing software professionally for 2 decades. I prefer Cursor.
Am I an outlier or do you just judge people for weird reasons? I’ve never seen an IDE person judge a terminal person, it’s always the other way around - what’s up with that?
I think it's more of a sign of a good engineer. I know a number of engineers that are good and don't really work with the terminal. On the other hand, every engineer I've worked with who was a 'terminal guy' was great. I think being good with the terminal is a signal that the person is willing to 'dig in' and understand stuff at a lower level.
> I think being good with the terminal is a signal that the person is willing to 'dig in' and understand stuff at a lower level.
You know what else is a good sign of them willing to dig in and understand stuff at a lower level ? If they dig in and understan stuff at a lower level. Let's judge people on what matters - the actual work and value they bring. Not status symbols like their IDE of choice or how fast they type into the keyboard.
Yes, that's why I said 'signal' and not 'sole metric used to determine worth'. Devs get so touchy about this subject. I didn't say being good with the terminal is all that matters, it's just an extra piece of information.
I’m very comfortable with the terminal, but let’s be honest. It’s not very good at certain apps. For example, copy/pasting long bits of code or strings into and out of Claude code is highly annoying. Line breaks in weird spots, because of the terminal, for example.
Anyways, I use cursor for a number of reasons:
1. I still want very quick access to the code in the editor. So I want the IDE.
2. Generally solid defaults. Auto-compaction, plan mode, etc, all work pretty well.
3. When I switched back to it from Claude code, it was genuinely faster at running Opus than Claude code. Claude code was grinding to a fucking halt every two minutes.
4. So annoying to search and view your chat history in Claude code. I’m a visual person. I also want all my repos loaded into a big workspace. Cursor also does that great out of the box.
5. I don’t have time to redo my terminal setup again to optimize it for Claude.
Tbh, I’m not aware of much that Claude code does that you can’t also do in cursor. At the end of the day, the agent loop and tools are not that different, and the model is identical.
The tool you use to prompt it is not the hard part. I just work faster when I have everything easily accessible in one spot, which was easier for me to accomplish with cursor than Claude. I found it just got out of my way.
Someone instructing AI through the terminal is a bit like an office worker with a tool belt. I don't think you can say anything about their coding ability until they are coding without AI. Even if thats in notepad.
I actually never mentioned anything about actually using the AI tools integrated into Cursor in my post.
I think I'd generalize my post more to say the more often somebody reaches for the terminal, in my anecdotal experience the more proficient they tend to be.
Totally disagree. I find people still using cursor or other IDE centric flows want to review the code and be more interactive. Claude Code and Codex push agent autonomy and speed. Sorry but they go off the rails too much.
This. I hate VSCode as an IDE and is the reason why I have not used Cursor. I wish Jetbrains actually had some brains to build a better coding agent inside their IDEs (which I think are one of the best out there), but for now Im stuck with codex/cc + Jetbrains IDE
IDE ships it's own compiler, debugger, usually some sort of container/container orchestrator and database tool. I.e. the things that separate hobby projects from production ones. VSCode has these things but they're extensions, not build into the editor
I am not trying to be pedantic, I am really just curious because I’ve never heard someone saying something like vscode isn’t an IDE. I am mostly curious because those of us that write languages like Ruby or python with no compiler that typically debug in the terminal or repl feel like we have an IDE and refer to it as such.
I also think your characterization about hobby/production fall short. My fastAPI and rails codebases anre definitely not a hobby.
Thanks for clarifying though. Interesting to hear your perspective.
> I also think your characterization about hobby/production fall short. My fastAPI and rails codebases anre definitely not a hobby.
That's not what I said?
What I said the was the line between hobby/production is use of the tools themselves. Doesn't matter if you use them via a terminal, VSCode extension, or an external app.
Line between editor/IDE is the former requires downloading extensions but the latter bundles them in the final application.
You're right, the meaning has become pedantic so I'd put forward a new term because the line between editor and IDE is blurred... but there's still a vast difference between editing on vscode vs IDEA or Goland.
ILE? Integrated Language Environment. The former caters to a specific language ecosystem. Text editors are polyglots by contrast.
If you call text editor that has terminal bolted on that’s not IDE. It is exactly the same as if you would open terminal on the side. Grepping/search with regex is also far from code navigation that is available in an IDE.
Debugging tools and code navigation in a proper IDE like Visual Studio or JetBrains Rider is totally different development experience than editor with terminal.
why do annoying engineers have such a weird fetish/superiority complex about the terminal. Its an inherently inferior UI. Theres absolutely nothing you can do in a terminal that you cant do in a GUI, and every TUI is just jumping through insane hoops to support functionality thats trivial in a GUI.
Why on earth would you want to look at a code diff in a terminal? Why on earth would you want to use weird bespoke keyboard shortcuts to navigate sub-agents in a TUI session. Why on earth would i not want the ability to open/edit a file manually in the tool im using to write code.
As someone who has spent the last 10+ years working in Tmux - but is entirely comfortable on Mac, Windows and Linux desktop environments - here are the key reasons why the terminal experience is superior for me.
- I work a lot with data - and streaming data through text tools is twitch fast. If someone has a question about data - before anybody else can log in to their superset, or analytics database, and try and work through the SQL queries or charts to get the answer - I've already jammed the data through awk and got an answer.
- As an SRE - I work with a lot of systems that have pretty rich APIs - so being able to send a request, get the answer back in json, dump it into jq, select the parts I care about - maybe -c to compress it and ripgrep a subset out - is just fast.
- I work in a lot of contexts with a lot of different systems, datacenters, applications - tmux lets me keep all of them cleanly organized in a separate windows and subpanes. I'll have 15-20 windows open per week, and maybe a 5-6 panes in each- keeping 100+ different contexts (and scroll backs, bash history) - all nicely organized is really useful.
- I'm also a systems guy - and there is no other way to dig into a system but the terminal - netstat, ps, dmesg, /proc - these are all components that have only one credible path to investigation and discovery. If you aren't super comfortable in the terminal - zero way to learn about this stuff.
- Working remotely - means ssh. So - once again - terminal.
The Focus on the terminal is that it's the best tool (and in some cases the only tool) for so many of these tasks - and by performing these tasks a lot - you learn about systems - so the people who spend a lot of time in the terminal tend to know a lot more about systems than people who don't.
It's a terminal emulator with a command prompt that also gives you access to LLMs. But the interesting bit for people who work on a lot of machines is that this also works when you ssh to any remote machine.
So you can bring your agents to any remote system, it even works somewhat well for network devices.
The thing you can do in the terminal that you can't do in a GUI, is glue together over 50 years of useful tools, no matter where you got them from or if the authors have ever heard of each other.
If your workflow fits entirely within a single app's GUI, then yeah, the terminal version of that app is not going to be as useful. But if that app doesn't exist yet, you can put together an 80% version of it for 20% of the work.
Historically, it's also a lot more resistant to rot. Brian Kernighan isn't going to start charging a subscription fee for AWK - and if he did, there are many forks and similar tools.
And, addressing a specific point - why would I want to view a code diff in a terminal? Sure, 'diff old.txt new.txt' is probably less useful than popping it open in a nice GUI with highlighting. But "diff old.txt new.txt | grep '^+'" will only show me added lines, or "| less" and type "/foobar" to jump to all mentions of foobar.
And this is like, the least you can do - the stuff you learn in the second class of "Using the Terminal 101". You can easily use this with git, as a building block to make a quick script to graph the number of changes over time in your repo. Yes, there's probably a GUI somewhere in your workflow that can show this (maybe you click around in Github to find it). But, maybe you also want to just filter that to changes in a specific module in the codebase, or an author, or quantify what module changed the most each month. If you've learnt the building blocks, the scriptability of the terminal lets you put that together quickly.
Anything you do in the terminal can trivially be scripted (automated). It's a self reinforcing loop of making life easier. After many years of working in the terminal, and making little scripts, my workflow is extremely fast, comfortable, and customized to me. You can do some of this via GUI tools, but terminal makes customization easy. Also, using keyboard shortcuts is just vastly superior to using the mouse, you can't convince me otherwise.
> Why on earth would i not want the ability to open/edit a file manually in the tool im using to write code.
I don't know what you mean by this. You can open any file in neovim at any time without leaving the program.
Being familiar with the terminal also makes building CI for the team trivial because I'm already familiar with how all the commands work in the CLI. I'm basically the goto 'devops' guy because I'm one of the few people who actually knows how to work in a Unix terminal.
I will say, TUI is not the same as CLI. I don't find a meaningful difference between a TUI and a GUI other than being able to use tmux or something for window management. I prefer gui tools for database management, querying, git diffs, email, all kinds of stuff.
As for the superiority complex. I've got no judgement on people who prefer the GUI. I have many excellent coworkers who primarily use GUI tools. Having said that, every engineer I've met that works primarily in the terminal has been great. It's a very strong signal of technical competence in my opinion, but terminal familiarity being a signal of competence, in no way makes GUI usage a signal of incompetence.
Absolutely nothing? That's certainly not true. My experience is that those who grew up learning the command line are so familiar with it that they excel at navigating those bespoke keystrokes more quickly than any GUI user who has to scroll, point, and click. Add to that features like command history and autocomplete, and shell users are often far more productive than GUI users.
Like any other computer UI: Terminal-based programs (whether ultimately windowed in a GUI or not) didn't start off being familiar. But for those who use them, they eventually become familiar.
And that familiarity transfers between different systems. Windows, Mac, Linux, whatever: The flow of any particular terminal-based program is the same everywhere that it can be used.
It's tidy, and light. It's also network-transparent, and things like ssh keep it secure. Multi-user support is the norm instead of the exception on systems where terminals are common. It doesn't interrupt anyone else's work like something like using Anydesk to access some GUI desktop somewhere else can.
The keyboard shortcuts are annoying at first, but they're faster than mousing around in a GUI -- and once learned, they're approximately impossible to forget.
(You're free to hate terminals if you wish. I don't care if you justify it; I'm not your boss.)
I did a lot of automation at my last job, which was closer to engineering in the classical sense than engineering in the software sense. The automation mostly amounted to web scraping and interacting with local systems, with a generous amount of logic in between. The largest roadblock wasn't the scraping or the logic, it was the inbuilt assumption of the local programs that there would be an ape with a mouse and keyboard driving the interaction. Outputs that could have been text needed to be copied to and read from the clipboard, or exported as a spreadsheet. Inputs that could have been text were only accepted in spreadsheet form. Pixels needed to be read from the screen to tell when one step was complete and the next could begin. Mouse clicks and keystrokes sent to and fro when it all could have been a series of commands. I cannot count how many written processes and procedures existed could have been a bash simple bash script.
I'm not arguing that these programs should not have a GUI, for that was the simplest way to use them, but the lack of command line functionality places a hard limit on the productivity of an organization, and ensures that the only progress on that front comes at the expensive of exceedingly limited developer time.
"But nobody knows how to use a terminal anymore," I hear you say. Well of course they don't, nobody under 35 without a background in programming has ever had cause to use one. We made everything so simple that nobody ever has to learn anything. That isn't to say that people cannot learn, just they have been robbed of the natural opportunity to do so. Otherwise intelligent people never progress beyond the manual step-by-step interaction that passes for "using a computer" today.
A computer is a tool in much the same way that the a machine shop is a tool: it is a tool that can build other tools. The role of software developers should not just be to build simple tools that can do one task in isolation, it should be to build tools that less technical people can use to build the tools they find themselves in need of. GUI-only programs are simply not fit for that purpose because they lack composability of simpler terminal-based programs.
Yah this judgment and arrogance is so annoying in tech. And worse it stops us from learning. Some of the best lessons of my career were when a new developer asked a question often taken for granted or we implemented a design pattern to make coding more approachable.
Terminal/CLI is superior if you know what you want to accomplish. There is no faster way of doing stuff on a computer if you know what you’re doing and having tab completion and knowing all the incantations.
GUI is superior when you have to figure out what to do and what needs to be done. GIT branches state checking, finding out a switch to enable a feature so much easier in graphical interfaces.
I find that what you call „annoying engineers” are just people who have their tools and use them efficiently … but once they have to step out of their path they become annoying or even obnoxious.
The opposite is random person who expects everything to be easy and available in less than 3 clicks — well good luck engineering an interface for complex system that does that, not everything is Facebook, there are systems where you need to spend time wrapping your head around.
Honestly the TUI in most of these coding agents is so fancy I have trouble thinking of them as "terminal". I use Pi Coding Agent and the fact that it's terminal means it's easy to run inside something properly sandboxed in a YOLO mode using normal bash commands instead of relying on individually sandboxed tools.
Once I got the tmux settings for proper scrolling and whatnot it feels fine. Honestly the TUI of tmux is the one that really enrages me - so much complexity for just "I want to switch terminals on my remote".
With the one caveat that it does not (yet) support copy/paste buffers or copying from the scroll back buffer. It’s also adamant that you use a mouse to select/copy - anything more complex and you need to pipe everything into neovim (or whatever text editor you use) and do your work there. I love zellij - it feels like the future - but hard to give up keyboard based select/copy and pulling from my scrollback.
why do annoying engineers has such a weird fetish/superiority complex about the terminal. Its an inherently inferior UI. Theres absolutely nothing you can do in a terminal that you cant do in a GUI, and every TUI is just jumping through insane hoops to support functionality thats trivial in a GUI. And guess what, you can just open a terminal in cursor! who knew!?
Why on earth would you want to look at a code diff in a terminal? Why on earth would you want to use weird bespoke keyboard shortcuts to navigate sub-agents in a TUI session.
Why should others spend their valuable time helping you? Especially when you insult the people you want to answer you "fetish/superiority complex" just demonstrates your own prejudice.
Personally I ask AI for a summary of positions, and prompt to provide some good articles on a subject - ideally articles from supporters of either side.
I actually was a Cursor advocate / CC hater (go back in my comment history), and now I use only TUI coding harnesses.
To start a big part is just the efficacy of them, which comes down to the model and the harness logic itself. CC is good, it's sub-agents, loops, background jobs / agents, skills/hooks/etc have typically been pretty far ahead though others are constantly catching up.
But you're sort of missing something. I use iTerm, so to me it's not the TUI itself, it's iTerm. And while it's imperfect, what I get is this:
I can open and close sessions nearly instantly and tile and window and tab them as flexibly as I want, plus it's a system I'm familiar with in terms of shortcuts etc. Has my configured theme, fonts, etc all set up. Every GUI app is different, every TUI app has half of the UI already incredibly familiar to me, it's not "just text", it's iTerm.
That also means they all are the same - I run Codex and Claude and pi side by side, and i switch between them with no overhead and minimal mental model shift. Sure, different harness does suck but that's the same issue with GUI just with an additional new layer to learn.
Smaller thing is because it's all text, there's no limits on my ability to copy things out. And it's a really fast text renderer that can render tens of thousands of rows efficiently. Many GUIs have various dialogs, unselectable areas, virtualization, or just slow past a point. I trust my terminal scales.
I don’t think mapping tooling to ability makes sense here, particularly when the “advanced” tools here just abstract more away, though I agree that Cursor is terrible. So many useless windows.
Opposite. The weakest engineers trust CC or codex, stopped reviewing the code and push slop PRs. Those still acting in the loop move faster with better architecture and coding patterns and aren't losing their skills.
>That said, I think we're in a narrow window of time right now where any of this matters. Prompt "engineering" and working around your tools will be over in a year or so.
As previously C system programmer, I barely use IDE, cli agent for coding and desktop codex for various non coding tasks.
Nowadays settled to codex/gpt5.5 and it does really good job
I didn't like cursor when it first came about but now I use it for my personal projects. The plan is good value for accessing different sota models occasionally for planning. Composer is actually really good, and fast.
I'm not sure if it's because my personal projects are small enough I know them inside out and my work project is huge, but I prefer terminal code agents for day work over ide integration.
i moved all my AI coding over to claude code when claude 4 came out, but kept cursor for the tab complete. since opus 4.5 i haven't really needed the tab complete so i canceled my cursor sub 6 months ago.
switched back to vscode so i'm not exposed to the potential mess that is openvsx too. trying to get used to zed but i'm just so used to vscode
Also using codex as a full time partner these days. What do you think happens in a year or so that changes the way it works around the tools? It becomes the only tool we interact with, and it assumes control over the others?
Could you clarify what you mean by “…will be over in a year or so”. Genuine question. Is it that models will be so good that none of this matters or we will need to go back to older ways?
I recently made an npm package with a small C helper that runs in the background. The JS/TS code is 99.9% unit test covered and for sure "cleaner" code. Just my opinion though.
I primarily use Cursor, compared against Claude and haven't used Codex before - to me the benefits:
- Composer 2.5 is cheap, fast, and very effective; I don't even use other models that much anymore, as it's usually marginally better for way more cost, though sometimes I do for specific things like making better translations for our app than a coding-specific model could normally output
- It makes setting up and maintaining Cloud Agents super easy, the agent can basically set up itself and if anything changes that makes it not update properly, then it tells you and can fix itself easily
- Easy to move agents from remotely in the cloud to local and vice versa
- Easy to work with agents via either comments on GitHub, or via the web app on my phone (though it is relatively constrained relative to the actual desktop UI, which is a bummer)
- Code reviews with Bugbot is surprisingly good now vs when it first came out, as is the Security Agent, while being an order of magnitude cheaper than stuff like Claude reviews
- Automations are easy to configure and manage - crons, in response to repo events, etc. For example I don't use Renovate or Dependabot much anymore since LLMs can update deps and investigate subtle breaking changes much better than a dumb version bump script can
- Limits are obvious rather than Anthropic's mysterious quota amount that they don't explain at all
- Queueing up messages rather than the agent taking in new messages mid-work and then trying to mesh them together somehow - I find queueing much more predictable and easier to work with
- Plan mode is good too, but not particularly different than any other agent
- Easy to jump into the Editor view and actually go in and manually code things or interactively code with the LLM when you need to, since sometimes LLMs just suck at doing certain things autonomously. I'm not in the "stop coding bro" camp, I still like to take the wheel fairly often.
Literally the opposite is true. Being a text editor at its core, and by spending a lot of effort refining the human-AI pair programming experience, Cursor makes sense for someone who wants to lead code development with AI being a speed multiplier and a team member.
A space company is buying an IDE for roughly the cost to build 150 of world's most expensive modern hospitals [1]. How is this in SpaceX's interest? Isn't it kinda bizarre that Elon is pivoting SpaceX to something else?
The company that just IPOed is already overwhelmingly "X AI" financially, regardless of the fact that it says "Space X" in the marketing. Whether SpaceX also buys Tesla is hardly even going to move the needle.
StarLink is a bad investment too in my opinion. It is fundamentally a US company that must charge US prices but in the US there are relatively few people who need that tech. Only in more rural areas that are sparsely populated do you not have access to fiber or cable which give better speeds and latency.
The majority of their revenue is overseas. But there they can’t reasonably charge $99/month. And there labor and land are cheap such that you can get fiber laid cheaply and quickly.
StarLink is the only thing about SpaceX that makes a profit. But that's only possible because no one else has a network up there hence it can ask for any price it want. Let's not forget that before StarLink there was Iridium and it got ousted because there was a better offer and I wouldn't be surprised if some other company will eventually go up with a better offer.
They aren't in the black at all, let alone "insanely" so. Maybe if they just did commercial launches on falcon 9, but instead they've been shoveling those "huge margins" straight back into r&d on starship etc.
The Falcon launch business is profitable on its own, which may be what GP meant. Once you include the ~$3 billion in Starship yearly R&D the space segment goes into the red.
It is what I meant. SpaceX has strong division between its operational and R&D arms, and Falcon launch has been insanely profitable from the first F9, even before reuse dropped their costs like a rock. They just invest those operating profits back into the company.
Starlink guided drones are tipping the scales for Ukraine right now. That's what enables the range required to successfully hit trains and fuel trucks supplying Crimea through the "land bridge" made from conquered lands in South Ukraine.
That plus satellite comms. Both Russia and Ukraine were very fond of how it improves communication.
Edit: meant to say that in the time of global instability, weapons tech is going to be valuable.
Not so much Ukraine using drones that use starlink, more the sheer number of starlink sats up and planned, the numbers in the proposed other constellations going up and planned, and the ongoing propensity of Chinese rockets to fragment after use rather than safely de orbit.
For interest:
An Orbital House of Cards: Frequent Megaconstellation Close Conjunctions (2025)
Our calculations show the CRASH Clock is currently 5.5 days, which suggests there is limited time to recover from a wide-spread disruptive event, such as a solar storm. This is in stark contrast to the pre-megaconstellation era: in 2018, the CRASH Clock was 164 days.
Yep, it feels like forever a niche market. As soon as you think about it like fundamental infrastructure (which as I understand it is what we’re supposed to think about) it quickly becomes obvious that cables are better suited. Wealthy enclaves, digital nomads etc will pay for it, but that’s not going to get the revenue they’re hoping for.
You mention that cables are better suited... but that is the whole point of Starlink. Getting cables everywhere is either not feasible or expensive. The rural properties are already mentioned as a very big source of usage right now around the world. But you also have everyone who is using it for some mobile purpose. Whether that be planes, boats, RV's, cars, military, etc... I've even seen people live streaming things using Starlink due to cell towers not giving good enough connections where they are.
While they may not be ideal or people living in Urban areas, they also aren't limited to selling to people who are physically connected by cable like a regular ISP would be.
I think you're underestimating the market here. Especially on the enterprise side when you start thinking of things like airliners, cruise ships, etc...
Agreed, starlink is obviously advantageous from the global (and extra-global) perspective. Running cables is expensive. But so is launching satellites, and it’s hard to see how you reach a tipping point where satellites gets cheaper than the standard boring cable networking that supports the terrestrial internet
Starship is trying to get launch cost down to $10 per kilogram to low earth orbit. At that cost it is not clear to me that cables are better except for short distance point to point backbones. One other issue with cables is regulation or monopoly.
There are two different questions. Are cables cheaper and is Starlink profitable in the long term?
Launching a starlink satellite costs $1.5-2M per satellite for the current design. Each satellite lasts around 5 years and they need 42k of them.
Do the math and you get $13B to $17B per year ignoring inflation and ignoring the massive ground infrastructure that is also needed (costing billions more per year).
Residential plans range from $55-175/month and will make up almost all subscriptions. Most people will get the $85 middle plan, but we'll round up to $100/mo or $1200/yr.
Breaking even with no profit requires 10-15 million subscribers.
As starlink claims to have around 12k subscribers and just 10.5k satellites, I'd guess the system is profitable.
The cable question is still compelling though. Rural fiber costs around $25k per mile. This means that an average $15B/yr buys 600,000 miles of installed fiber every single year.
My conclusion is that the ideal solution would be Starlink keeping total satellites closer to the current 10k number ($4-5B per year) and spending the other $10B/yr running fiber to new areas.
Their subscriber base would drop as new fiber got installed in rural areas overlooked as "not profitable enough", but stabilize at still profitable enough from travelers and truly inaccessible places.
How many people actually live in areas that do not have some sort of broadband? That’s the whole point: this is a product designed for relatively few people.
I think there's a strong chance this is a case of it creating a whole new market though.
There are people whose current behaviour/situations will happen to benefit from this, and that may be a niche, but seems like there's a really solid chance many more people actually will change their behaviour in response to this being available. That's how disruption happens.
To be honest I can easily see the default changing if the service is good enough. I mean it seems like you basically get most of what is good about wired plus a whole load of extra previously totally unavailable benefits. For a price, to be sure, but that'll come down.
> labor and land are cheap such that you can get fiber laid cheaply and quickly.
Which is why all of sub-saharan Africa is crisscrossed in fiber, and every hut in those villages has an ONT zip-tied to the straw walls where they get their symmetrical gigabit service.
There is a huge market even just in the US for Starlink. The worldwide market for people who need internet access in remote places is positively gigantic.
Additionally, Starlink could be hugely profitable if they exclusively sold access to ships and aircraft. You'll never be able to run fiber to those.
They can indeed reasonably charge $99/month to many many millions worldwide who don't have any options for low latency, high speed internet access. You vastly underestimate the need for the service.
On the one hand you have a technology product that’s only relevant to rural consumers. Nine out of ten people have cable already.
On the other hand your margins are amazing because all you do is fly little boxes over everyone’s heads launched with government subsidized rockets. No linemen or plant-hire or contractors to sap your profits.
The biggest threat would be commoditized terrestrial wifi / 5G. The more cell service competition there is, the smaller the market for satellite, until it’s only applicable to 1% of the population (and the poorest 1% at that.)
More like 8.2 out of ten. Either way the remainder is still a pretty decent market. And that's just talking about people in the USA. About 75% of Starlink subscribers are outside the USA.
> and the poorest 1% at that
Not by a lot. People who live in remote areas in the USA tend to have much less money overall, but they tend to spend much less money overall, leading to a similar amount of buying power. Someone who lives remote is more likely to own their home outright or have a relatively small mortgage. Their socio-economic status can appear numerically depressed because the numbers generally don't account for non-monetary consumption. (You got paid a salary and bought salmon from the supermarket. Remote dude fishes for salmon in a local stream. You both traded your time for salmon, but remote dude's salmon is invisible to GDP statistics.)
And furthermore, for them, Starlink would be budgeted for like an essential service rather than a luxury convenience.
> Starlink would be budgeted for like an essential service
As someone who has lived and worked remotely for the bulk of time since 1960 or so, it's not essential ipso facto; myself and most of the people I know have somehow managed to survive sans this supposedly essential service for 60 odd years (since 1935 in my fathers case, he's not dead yet).
Its more compelling use case is a relatively cheap way to integrate vehicle GIS data across four to ten thousand hectares or so (ten thousand to twenty five thousand acres) for farming, mining, exploration, etc.
Globally, its not especially attractive for non civil applications (military use, etc) as it creates a reliance that can have a plug pulled at the worst moments.
You are correct that it's not something any military can/should rely on for any future conflicts.
That doesn't make it useless though. Ukraine certainly finds Starlink attractive for military use. Despite all the misleading headlines and de-contextualised quotes, SpaceX has been reliably on Ukraine's side of the conflict and has been an essential communications fabric for both military and civilian.
Just throwing it out there, Im a Verizon customer in the Tri-State Area and I frequently lose service on the train or bus on my way to the city. It's more than just a 'rural issue'. I looked into getting a StarLink plan for my commute until I realized logistically, at best, I'm a freak sitting next to the window with a satellite dish on the train.
Yeah i'm no fan of Elon but starlink is actually quite useful. It solves a real problem - I need network access anywhere on the globe, for any reason, at any time. they can easily charge what they wish for that (military, air travel, desert tourists etc.)
Most of the areas without good internet barely take home $100/m and a lot of them take home less than that.
If you hooked up every single cargo ship on the planet to starlink, you'd only add around 100k connections and average wages on most of those vessels is $5-8/hr (very few US/EU sailors these days) for a handful of people which tells you how much businesses actually care about their workers who do dangerous jobs.
Elon's award is tied to growing Tesla's market cap - it's pretty transparent that he's just trying to ball-of-mud together everything he can to hit that target and grab the bag.
What would it mean if SpaceX buys Tesla though? Does the combined market cap count? That would be wrong. Tesla buying SpaceX just for hist bonus and then rebranding to X would be classic Musk.
It's a game for him, but so ridiculous. While Tesla was pushing electrification and SpaceX pushing rapid rocket re-use I kind of tolerated Elon's antics, but since he got involved in politics and DOGE I can't bear it anymore.
>What would it mean if SpaceX buys Tesla though? Does the combined market cap count? That would be wrong.
I took a look at the proxy statement as they have it outlined in this [0] document (Proposal 4). As currently formulated, Musk has 12 operational goals (like ship 1 million robots or hit 50 billion EBITDA) and 12 market cap goals. These need to be paired together for the shares to vest; so, if he reaches the first market cap number, he also needs to fulfill an operating goal for the first set of shares to be earned. These earned shares vest in ~5 years.
This comes with a huge caveat. If Tesla changes control, those operational goals go out the window and his stock award is based solely on market cap (along with the shares immediately vesting). The share price for this is the greater of:
1. The last traded Tesla price prior to the acquisition, or
2. The per share price outlined in the acquisition.
The "easiest" way to take advantage is to IPO SpaceX (which he did), pump up the market cap, acquire Tesla for a sizable premium, and vest as much of the stock award as you can. It means you get to avoid the operational goals entirely and vest a bunch of Tesla (but soon to be SpaceX) stock.
My understanding is that Musk doesn't have a controlling share of Tesla. So by the letter of the law, a SpaceX acquisition would qualify as a change in control (because we've gone to a new majority owner).
The thing is, even the center of the road academics who refrain from such comparisons are noting the similarity. The MAGA movement meets the traditional defining characteristics of fascio. We can be rigid and say that only Mussolini‘s political movement can be properly be called fascist. But if we accept calling Hitler and the Nazis fascists as well, then we open the door to any movement that meets the criteria. And Trump does fit the bill.
You are conflating a strong regime with authoritarian tendencies with fascism. The link you gave could be used to define China and India as "Fascist" as well. Which they aren't. Besides, other categories exist to define regimes, but I guess now the public debate is binary: either you are fascist or you aren't. I agree that it reduces the amount of thinking required, which is always nice I guess.
In the case of Trump, caesarism is likely a better term. Unlike fascists leader such as Hitler or Mussolini, Trump doesn't have an ideology, is heavily corrupt and self-serving and is deeply influenced by a foreign power and lobby. Interestingly, he was stabbed in the back by his closest collaborators at the end of his first term, when he tried his luck. Like Caesar.
But Caesarism is less frightening that "fascism", so...let's ignore it! This strategy is a classic of the communist way of doing propaganda: Russians still whine about "Ukrainian fascists", and in the 80´s in the communist block, everyday you could hear complains about "fascists" at the radio. The Berlin wall is originally named the "Anti-Fascist Protection Rampart".
Why are you reusing soviet talking points, exactly?
Caesarism/kleptocracy is definitely a more accurate description.
People usually attribute the term fascist to this admin, because of the racial/ethnic retaliatory actions or the neonazi/neofascist base they need to overpower/outweigh the more moderate right.
Actually, Caesar was very self-interested and got very rich from his campaign in Gaul. Plutarch estimated that Caesar brought back (and sold) a million slaves.
The strong minority base that has been artificially inflated to advance the personal interests of the leadership has linear roots dating back to the XX century rhetoric promoters and activists. In Europe there's the militant and militia groups of Gladio and the likes that thrived in illegal affairs and that have persisted and were integrated into power following 90s (case in point in Italy Berlusconi basically legalizing the MSI, AN, FdI and Lega by using his privately owned televisions and media, to acquire an election victory).
Italy is a good example why it's foolish to use "fascist" to describe everything one doesn't like. The European politically correct press was wary of the new "post-fascist" leader of Italy, promising us black shirts and leather boots again. In reality, Italians got a classic right-wing government, which increased work visas for non-european migrants (as they usually do, since doing the opposite of what you were elected for is a common habit in Europe).
In reality "fascist" is a qualifier to frighten white people voting for a right-wing government. Trump base by the way is not fascist, it's the evangelical bible belt. Which is crumbling at the moment, by the way.
I assume that you read the link I provided. It's structured as a 10-point list. The 10 points of comparison are:
1. a mythic past and national rebirth
2. victimhood and humiliation
3. hierarchy and dehumanization
4. contempt for weakness
5. the cult of action
6. the leader as savior
7. the purification of institutions
8. propaganda and the assault on truth
9. the merger of state and corporate power
10. violence and terror
It's not stretching things in the slightest to say that MAGA and Trump fit these to a T, certainly post-1/6 if not before. If it walks like a duck, and quacks like a duck, we shouldn't be too timid about using the word "duck." Do China under Xi and India under Modi also fit the bill? Maybe. I don't think they check 10/10 boxes perfectly though.
I know we're running up against HN rules, so I'm trying to keep this meta and not object-level. I'm not debating whether any of this is defensible or good or bad policy or whatever. Just whether we should use the term.
The first Trump administration was arguably Caesarism. The second has proven to be an entirely different character. Just picking on one thread that is more on-topic for HN: the merger of state and corporate power (#9 on the list above) is a defining economic characteristic of fascism that we are only now seeing this term, with direct acquisition of large stakes in Intel, IBM, GlobalFoundaries, various rare-earth metals companies, and Westinghouse. OpenAI is in negotiations to do this, and even before concluding a deal the government has already been weaponized against their chief competitor, Anthropic.
Other things have progressed (regressed?) significantly in this second term against the above metrics. Unlike the first term, which mostly rode out the instability, we saw in 2025 far reaching restructuring of the civil service. And trans oppression and scapegoating, for example, is seriously reaching levels comparable to Nazi jew hatred and victimhood. As someone who lost my European relatives to the holocaust, I do not make that comparison lightly.
Seeing as ICE is currently building / converting hundreds of new detention centers all over the country, it is reasonable to be worried.
The definition you are using is far too broad, so you get a barnum effect where every authoritarian political movement could be shoehorned as "fascist". Ask real fascists what they think of Kash Patel.
Benito Mussolini, who first used the term fasci in its modern meaning to describe his party and movement, was not a racist. A cultural nationalist, yes, but he thought the nazi biological race ideology was bullshit and openly said so. The racial discrimination laws were only enacted after he went to Munich, hat-in-hand.
Yeah, anyone who openly calls themselves a fascist are probably white supremacist nazis, and I presume would take a dim view of Kash Patel for that reason alone. But genetics-driven racism (as opposed to cultural nationalism) was never a core tenant of fascism. Fascism requires an us-vs-them othering world view, but it doesn't have to be race-based.
Us-vs-them worldview is I think a very common thing in the world. Was the Soviet Union fascist? Is India fascist due to its extreme antagonization of Pakistan (and vis-versa).
What you say about Mussolini is false by the way and Italy enforced strict racial discrimination in Ethiopia. Political racism was commonplace everywhere at the time and is not very specific to Fascism. Unless 1930´s USA was a fascist country, which I doubt.
The existence of a fascist movement in the country doesn't make it fascist. France had a similar movement at the same time and wasn't fascist, for example.
Fascists don't support free speech. Please, learn what that word means[0] and then use it responsibly. We're already seeing the effects of misusing and overusing several other important words for political/societal discourse, and this is one of them. It's entirely unhelpful and lowers the quality of conversation here.
[0] The Doctrine of Fascism is available, for free, in several places. Start there.
One doesn't need to be a free speech absolutist to not be a fascist, far from it; and the exceptions prove the rule - those examples are the definition of exceptional. "10 times…" I can remember people getting banned en masse for changing their profile pic to an NPC character. The difference is night and day.
Have you heard of the term "bad faith argument"? That's a good example.
> - Opposition to political and cultural liberalism
> Yes, banned speech on twitter that trans people use.
I had to check this one. I found two stories. [1] shows that people were warned about speech (e.g. using the term cisgender) but were still able to post without problems. [2] showed that X's rules were updated to prevent harassment in the form of deadnaming or misgendering. I doubt that either supports your assertion to be true.
As to the rest of your answers, you seem to have given away that you didn't read the Britannica article at all:
> - Education as character building
> what
You should know, right? You read the article and made the accusation. You've done this several times:
> - The leadership principle
> idk what this is
> - The “new man”
> or this
There's not really a need for me to be pedantic in the face of such slipshod work, but then again, perhaps you think that not being slipshod is being pedantic. Which would also be wrong, but telling.
Try learning about what fascism is, that would be my suggestion. With a book, while putting your phone away.
pretty much exactly what i expected, you ignored all the YESs and called me dumb and made a ton of assumptions. You dont have to fit a definition 100% you're going to hang on to all the tiny little details and say wholey that he's not a fascists. I think he hit's the major point of the definition. Pedantic bootlicker. The reason I didn't pick up a book at your request is because I'm not wasting my time with you. And, no I didn't read the article, you did and linked it. I only read the definition.
Fascists do claim to support free speech while wielding the apparatus of state to ensure that any speech that they disagree with is suppressed, eg on college campuses.
> Is Musk wielding the apparatus of state to ensure that any speech that he disagrees with is suppressed
He does not need to do that, he has his army of people like you that go around on all socials and wherever you can to post praise, defend him and attack any critics as he was some sort of uber genius savior of mankind. The Musk cult is real.
So... are you arguing that Musk is an opponent of 'free speech' because... a bunch of private citizens argue with you when you speak critically of Musk?
The whole thread (and to be fair, Musk himself most of the time) has lost the plot, since the phrase "freedom of speech" is about the government not restraining or punishing speech. Honestly, what Twitter allowed or not before or after Musk is entirely irrelevant to the concept of free speech. Twitter/X is still just some guy's website.
Musk explicitly centered his DOGE effort on rooting out “woke” programs. This included pulling funding from independent research under the auspices of various government departments.
The Trump administration spent considerable effort strong arming colleges on account of speech that they did not like. Even if you subscribe to FIRE’s “Chicago Statement”, you should find the government interfering with colleges like this a break-glass scenario. “Individual Rights in Education” surely include the right to hold beliefs and teach in a way that the current administration is uncomfortable with.
So yes, these are fascist activities wholly consistent with the other fascist activities that MAGA movement is engaged in.
> Be kind. Don't be snarky. Converse curiously; don't cross-examine. Edit out swipes.
> Please don't post insinuations about astroturfing, shilling, brigading, foreign agents, and the like. It degrades discussion and is usually mistaken. If you're worried about abuse, email hn@ycombinator.com and we'll look at the data.
I feel like it has to eventually become X. Just because he’d find it funny for the stupidest people in the world to post about how he’s turned Twitter into a $2t or whatever company.
> Does the combined market cap count? That would be wrong.
It counts and it's not wrong.
Sure, the Tesla award takes into account any M&A but growing a 2T company to 3T is a 50% increase.
While growing a 1T company to 2T is a 100% increase so it's expected to be easier for him to hit his award targets with the companies merged as opposed to not merged.
If we combine the market cap of the entire S&P500 we get close to 70T. That doesn't mean any of those individual companies are any more valuable or any of the investors are any richer. It makes no sense that Tesla shareholders would be ok with paying out a performance bonus just for M&A that doesn't grow the value of Tesla and would just dilute their shares.
Well yeah it doesn’t make a ton of sense but the shareholders already cast their vote on this. I’m not sure what they expect would happen but I also don’t think they care. Elons shareholders are a special kind of financial sycophant.
We're talking about his Tesla grant so SpaceX's grant is irrelevant.
The text written in the link you did not read, states "Notwithstanding Sections I, II and III above, in the event of a Change in Control, the Operational Milestones shall be disregarded".
Because that's the grand plan from the start. You don't put a sentence like "In other words, upon a change in control where Tesla is acquired, vesting of milestones under the CEO Performance Award does not require the achievement of a matching Operational Milestone." if you didn't think Tesla could be acquired.
The goal is to give Elon his 2012 award package but making it his 2018 one.
No. He is making it an AI company. The prospectus makes that clear. Everything is in service of training and deploying AI. Twitter is data, distribution and marketing, space x is distribution with data centers and internet in space. Cursor is training data and hostile distillation.
That prospectus didn't make anything clear. It was pictures of rockets and rubbish about "the light of consciousness". The only real information was buried deep in the middle and it showed a company with poor finances and no clear path to success. Certainly not something worth the likes of Amazon.
How much mass do we need to put in orbit to radiate 10 GW of data center heat? How much rocket fuel does it take to lift the rocket fuel required to lift that mass? Which part of this is cheaper or more effective than building data centers 100 feet below the ice in Antarctica ? Other than serving workloads that are already in orbit, I don't see the point. How much addressable market do we have for AI compute that is already in orbit?
The appeal is solar power efficiency gains. And yes it is a notable amount of extra mass, but not a deal-breaking amount. Latency is pretty minimal regardless of location (sub-100ms) which is not nothing but it's minor for things like serving AI inference.
Not saying I think it's the best idea, but it is theoretically feasible.
When you say it is not deal-breaking - have you looked at any math? Real world data finds we need about 6kg of radiator surface to radiate 1kW. [0] . A 100MW datacenter would need about 600,000 KG just of radiator mass. ISS was only 450,000KG and it took 13 years to assemble. Meanwhile, we're projected to launch 600X that data center capacity on earth in just the next three years.
To say that this is pure folly is an understatement. It is not folly, and it is not foolishness. It is fraud.
It exists, but "works quite well" is only for a small amount of heat. It takes a big radiator to get rid of heat to ax vacuum. Much easier and cheaper to just keep the computers on earth and cool them with standard air or liquid cooling setups.
The required radiator for cooling isn’t that much larger than the required solar panel for powering the thing in the first place, and you don’t see everyone saying those are impossible. Is it easier to keep them on the ground? Obviously, but that wasn’t the claim. It just isn’t nearly as hard to cool things in space as a lot of people seem to think.
NewBird AI is an AI company which had an extremely irrational stock price bump because it's all about AI. It AI'd its AI with some more AI to AI harder with AI... and that equaled money.
If you're writing a prospectus right now and want a lot of large institutional investment you put AI between every letter. We are in a bubble, I don't think there's any disagreement about that, when the bubble will pop nobody knows - but while we're in this bubble everything is AI. I think it's unwise to read that prospectus in good faith given all the other factors (the court case against OpenAI, the race to IPO first, Google's additional stock grant, Elon's history of corporate bailouts) that are pretty plain to see.
To launch the data center into space and the satellites that deliver the data to users over star link. The same reason standard oil owned railways. Controlling distribution is a competitive moat no one else can replicate. Only antitrust stopped the standard oil monopoly. Not market forces.
Counterpoint: what if he lied in the prospectus? With Elons proven track record of misleading investors and customers I think that possibility has to be considered.
I agree with this, but it seems so crazy to me. How can money be a motivator when you're that rich. I'm not even "rich" but I'm already at a point where money is far from my #1 motivator.
I LOVE puppies, but if I had a trillion of them the last thing I'd want is another puppy.
An interview I recently read with Seth Rogen was very illuminating (from the NY Times):
"You know how every once in a while you read one sentence and it snaps your whole perspective into place? I remember reading that book “Going Clear,” about Scientology, and there was one sentence about how if famous people aren’t treated in a certain way, it makes them think they’re not as talented as they wish they were. Like, if I go to a restaurant and I have to wait 20 minutes for a table instead of them just seating me right away, am I not as talented as I thought I was? If someone has a nicer hotel room than me on the press tour, does that mean I’m not as good an actor as I thought I was? I think that’s how a lot of famous people interpret how they’re treated."
I think the same applies to Musk. The money is a proxy for how much everybody thinks he is a special genius. Anything in his life that makes him feel less special requires more validation that he is, and money is the easiest validation he is able to acquire.
Did he really win? His kids all hate him, his wives all left him, he has no real friends. Anyone he interacts with is purely after money and power they can get from him. He spends all day being angry on twitter.com.
I think the point of the Rogen quote is that there's no real way to win. This is not cope by the not mega-rich and famous. It's an inevitable result of wanting universal acclaim.
Musk has more money than most of us would dream of, but the game isn't over until it's over.
Speaking just for myself, I've lost respect from Elon Musk. I admire Musk's accomplishments, especially Starship and the Falcon rockets. But I don't respect Musk's personal judgement, his moral integrity or his ethics.
He doesn't know me, and he doesn't care about my opinion (or care about ethics for that matter). But there are a lot of people like me who used to respect him and no longer do. He's surrounded himself with fawning sycophants. At some level he's got to know this, and that the people pretending to pay him respect aren't themselves worthy of respect.
Musk is probably a clinical narcissist (NPD). If that's the case, no amount of power, status, or riches (or ketamine) will ever be enough
Their motivations are often cartoonishly superficial and... well... stupid. Stupid in ways that are baffling to most people (even most other neurodivergent). The kind of stupid that drives somebody to secretly pay pro-gamers to play games for them so they can pretend to be a pro-level gamer, only to then expose their own fraud by playing the game themselves on a live stream, without knowing how to actually play it. And then pretending to have connection issues when people start noticing.
I have no trouble believing Musk has simply internalized the identity of being the world's richest man and now has a pathological need to maintain that status, no matter what
It can't and it's just not. People use the word "money" for different things. He's not doing it for another bill or a number on some screens- neither are most employees of those companies. That's just projecting values on someone else.
The things they're trying to accomplish require extreme amounts of capital.
> The things they're trying to accomplish require extreme amounts of capital.
This could really cut either way. Like assume you mean he’s trying to do something good or noble (lmao) but the other obvious way to read this would be that he’s interested in becoming wealthy/powerful enough to bend the will of a nation before him and burn the world for a laugh.
It’s just funny how the markets seem to want to value him as some sort of AI visionary when his companies are not even in the top 5 for AI, despite his endless resources.
The way I see it, he's already well past the "wealthy/powerful enough to bend the will of a nation before him and burn the world for a laugh" but still needs to keep working to do his (and companies) stated "good or noble" goals. Why do the fervid anti-Elon people not even bring up his many flaws or mistakes and always concoct weird reality-orthogonal situations?
Ok sure I can be more more direct: how does aligning himself to the republican party and taking a detour to unwind as many social programs as possible in any way help his stated goals, corporate or otherwise?
The only answer is the dude is a self-service piece of shit. He has no principles and is not trying to accomplish anything other that exploit for the sake of exploitation.
That's not the only answer, it's just the one you want. There was a clear coup of the Democratic Party (or just shifting of values) that was jeopardizing both the actual climate change issues and the mission to mars. You can keep trying to find ways to justify the idea that he's just completely selfish, just wants money, or whatever, but you're going to have to do a lot more work.
I don't think you can describe his beliefs using booleans like that: you have to use a numeric scale. It would be be correct to say: Elon would need a hell of a lot of ketamine to believe a colony on mars is a possibility.
It's not exactly like Cookie Clicker. Many people definitely like seeing the number going up, but for most people that get to that point of wealth, the goal is power that the money represents. A human being may struggle to relate to them, but they really are motivated by the sole desire to own and control everything.
Yes you can. Money is a medium of exchange. Stock is a medium of exchange. But they buy the ability to effectuate your will in the world. If you have a vision of the world money, human capital, political power are not the measure of success or the goal. They are the instruments of executing a vision.
Some people aren't motivated by money, they are motivated by reputation, or pathology.
Yep, bundle it all together, rename the whole company X (for added confusion), get a few final pay outs in the hundred of billions if not trillions and then Elon flees to Argentina as the whole house of cards crumbles.
Does anyone really believe he’s doing all of this just to sell out and live on a beach somewhere? This dude sleeps at his own factories. He literally works like 24/7 and has no personal life. If this was all a cash grab he’s had dozens of opportunities to cut and run with well beyond F U money. If you wanted to scam people there are a lot easier ways to do it than repeatedly founding revolutionary technology companies.
I’m not denying that his companies are awash in zany financials, but I don’t think that’s ever been the point
The optimistic take is that he is doing all this to achieve his space goals, that the ends are justification of the means.
But when he is interested in actively interfering with elections the world over and probably spends more time on twitter than anything else, I am not so sure.
The thing with the sell out routine is that they think it is all working until it suddenly isn't. The money isn't about living the high life it is about having the high score. Bezos is probably trying to figure out how he can take the No 1 spot right now. If Elon is smart, we will cash out right before the fall.
They also seem to be desperate to buy their way into a monopoly, even though the company itself has a long track record of failing to deliver anything noteworthy.
FOMO; Apple grew to one of the biggest companies over time and people were like "h*ck I should've bought it years ago". Tesla then seemed somewhat similar - also 5 years ahead of the curve for EVs, even if they reinvented things that didn't need to and their construction was substandard, so one group of investors bought and boosted the stock to impossible heights. Now SpaceX is the next one, and people want to get on board.
I don't think anybody actually believes SpaceX will be worth it. I think some still believed Tesla would be, but their competition has caught up. Everybody is just there to ride the wave of mass hype and FOMO, thinking everybody else is an idiot.
This has been coming in a few waves over the years in Tesla, cryptocurrencies, the Everything Bubble, etc.
Yup - given that it’s $60bn of stock, now is an excellent time to do it, as the current valuation isn’t even irrational. And I say that as someone who believes they have great long term prospects.
> It’s all in a stock that may very well be near its zenith when this closes
No, it is not. This is not legal or financial advice, but I believe the stock could easily rise 2.5x - not because of its current or future financial condition, but because it is run by what may be the most skilled fraudster our planet has ever breed. Charles Ponzi himself couldn't have pulled this off. While most CEOs are careful about what they publicly say or "predict," Musk's companies are fueled by increasingly fantastical projections. The consequences have amounted to some $1.5 million in SEC fines that, relative to the value created, were negligible - and that was under the previous administration. The current one won't be any more aggressive. The closest comparison I can think of is Trevor Milton and his famous "electric" truck that was filmed rolling downhill under its own momentum. He went to prison for that, although he was later pardoned by Trump. Many people have lost fortunes betting against Tesla based on fundamentals. SpaceX's shareholder list includes so many influential and powerful names, including people closely connected to the current administration, that I find it hard to imagine the stock being allowed to fail in any meaningful percentage. Obviously I'm exaggerating when I say the government would send agents door-to-door to collect valuables from American households to plug any hole in the balance sheet before allowing the stock to fall significantly. But that's honestly closer to how protected I think the company is than what traditional financial analysis would suggest. I'm nobody special, just someone with about $1.8 million in a stock portfolio. Yet this thing called SpaceX stock gives ordinary investors like me a chance to ride alongside the biggest players on their way to even larger fortunes. They are guaranteed not to lose money, and to me its not personal.
I mostly agree but it still seems like a huge risk to play along with the memestock-ery. I have no idea how long this can be sustained. Like, would the upcoming elections going a certain way cause things to unravel? Or, say, the Anthropic IPO tanking and shaking investor confidence (highly doubtful, but a relevant example.) Or would some other unrelated event essentially cause a gigantic rug-pull?
And this is before getting to worrying about what kind of large-scale market manipulations we're up against (e.g. relevant example: https://www.forbes.com/sites/hershshefrin/2025/04/05/signifi...) Compounding this is the fear that even the indexes seem to be compromised (fast-tracking SpaceX etc.) I'm paranoid but at times the entire US stock market looks like a gigantic pump-and-dump. (Seriously, some of the stocks, like NET, rise and fall with highly predictable regularity.)
I can't tell, and I know I'm not nearly smart enough, attentive enough, and definitely not well-connected enough to have the perfect timing required to survive a sudden shift or whatever game it is the big players are playing, so I'm just trying to play it safe (which is also hard, because everything is impacted at this scale!)
The total addressable market (TAM) for SpaceX is finite. There are only so many nation-states and large corporations that want to launch payloads into orbit.
And even if their internet service provider is uniquely capable for now, it only fills a strategic need for certain customers.
So instead, Musk and Co. need to find bubbling market trends that look like they will have huge gigantic TAMs to justify the potential growth of this company.
All markets are finite. But you're thinking too finitely -- remember that there was a proposal to use Starship (BFS?) as a point-to-point method of people transport too (London to Sydney in under 50 minutes I seem to remember).
You also have other services: Starlink is an obvious one they're pursuing now, but there's many other things that they could branch into with no effective competition right now, from harvesting resources such as Helium-3 to Rare Earths (ironic name), to... (thinks for several minutes) banishing people to the Phantom Zone?
But you get what I mean, it's not just about rockets, it's about the things cheap and reliable rocketry enables.
Musk is a genius creating really exciting ideas. No doubt about that.
But as they say,"the devil is in the details"
- Can Starship transport people from London to Sydney safely economically, compared to Boom, which is working on a supersonic passenger aircraft ?
-Why can the boring machine dig tunnel at much lower cost than it's competitors? Maybe it's because the everyone else tries to dig tunnels for trains, which have a much larger diameter than Musk's boring machine, which only fits his "Teslas at a tunnel" concept?
And it might be a good idea. Worth a try. But be honest about it.
-Sure, data centers in space probably have some great uses, and I'm happy he's trying, but will they ever be more economical than deploying servers on the ocean? On countries with very cool climate?, powered by new energy technologies?
I'm not sure I understand. From a quick google search it seems like they are designing and manufacturing it themselves (with a few partners). Am I missing something, or is your information out of date?
The industrial bases of entire nations military aviation have been taxed for years/decades to produce supersonic capable engines simply for high maintenance fighters, now do that for affordable maintenance low headcount passenger jets. I'm not going to say impossible, but it will take more than VC money and anonymous 'partners' to convince me its likely.
For it to have a chance the US gov. needs to do IP transfer and pair them with NASA for test as they did with SpaceX, more realistically simply twist Pratt&Whitneys arm a bit...
Boom is planning to operate at Mach 1.7 (approx. twice that of current aircraft) with a range of about 4250 nautical miles (4900 miles, 7900 km). Over land they're only going to fly at Mach 1.3 to reduce sonic boom effects.
That's not enough range to do London - Sydney, it's not enough to do Los Angeles to Sydney either, or Los Angeles to Tokyo, it's basically a replacement for trans-Atlantic flights only because even US cross-continental Mach 1.3 is only about 50% faster than a 737 (3h vs. 5h)... It's pretty much geared to the prestige market only.
Under an hour to anywhere on the planet, meanwhile, is absolutely someone would be a premium for -- and they'll do it for most/all long haul flights if it was available.
Starship isn't a proven manned transport vessel yet. And Boom was founded in 2014.
If I was to hazard a guess, the BFS concept is 2035 at the earliest, most likely 2040 -- and it'll be at a much smaller scale than previously advertised.
Which hasn't yet been proven to be either technically or economically viable, even on paper. It's a pipe dream.
The cynical viewpoint is that this is Elon capitalizing on current datacenter hype to inflate SpaceX's valuation based on theoretically overcoming tremendous amounts of hard physics problems, over the next 5-10 years. As he did with FSD, Boring Company / Hyperloop, Twitter, etc.
Neither was reuse of rockets and I remember the ex-boss of Arianespace laughing at those bozos of SpaceX who try to pretend that they are a serious space business.
Musk made some bad bets, but also some good ones (Falcon rockets, Starlink) and some at least promising ones (Starship, Neuralink). And Twitter bought him enormous political influence - I wouldn't consider this a failure either, from the realistically-cynical point of view. That cannot be measured by revenue alone.
I think the jury is still out on the impact of all the other social networks.
If the connection between fast falling birthrates and smartphone addiction is proven, the total global loss of life (in this case, never born life) due to the products that many of us here helped craft into perfection may rival that of Mongol expansion under Genghis Khan, and Twitter/X is hardly the worst offender in that group. Even Twitter's impact on the balance of left/right politics in the world is relatively transient and small when judged against this horrible development whose aftereffect will stalk the world for a century.
Yes, it wasn't an actual intent of people like Zuckerberg, but as far as catastrophic failures of civilizations go, they don't have to be intended.
Then there is the role of Starlink terminals in an actual war.
The losses caused to the Russians by drones using Starlink for connection are pretty painful, and when SpaceX switched off non-whitelisted terminals in the theatre of war, the Russian army was thrown into disarray due to sudden failures of communication between their units. AFAIK they haven't yet fully overcome that problem.
The Russians certainly wish to have something like Starlink right now, in 2026.
In the first weeks of the war, Russia targeted communications infrastructure in Ukraine. They performed cyberattacks against Viasat, rendering thousands of satellite modems unusable. They performed destructive cyberattacks against Ukraine government and telecommunications targets. Cell towers and fibre optic lines were targeted, as well as electricity infrastructure. And they started aggressively jamming wireless communications. Much of this was disastrously effective.
Starlink was one of the only reliable ways Ukraine had for remote military communications, without which Ukraine would have not been able to defend its territory nearly as effectively. Though it's impossible to know, it's plausible that Ukraine might not exist today but for Starlink.
Personally, I don't think so, but let us at least try to be consistent.
"If an unpopular person's corporation C succeeded at activity X, it is the success of the regular employees and everyone but him, but if his another corporation D failed at activity Y, it is solely his responsibility and shame (if not a proof of outright fraud)" is a classical emotionally charged double standard.
Do you think there's a single Apple fanboy who thinks that Steve Jobs ever had any novel insight in programming, or had any novel insight in circuit design?
Now, with that answer in mind, allow me to contend that there isn't a single Elon fanboy who thinks that Elon is personally inventing automotive and rocketry technologies out of whole cloth.
What both men did well is identify promising unconventional technology pathways and steer capital investment towards them. Jobs had a knack for understanding computers as a consumer product, and for communicating the value of new products. Musk has a knack for understanding the limits of physical engineering, and the wealth (and appetite for risk) to spam the right "build" buttons endlessly.
Beyond a narrow range of remarkable competencies, neither are particularly interesting persons. I wouldn’t look to either of them for takes on sociology, politics, biological sciences, philosophy or chord progressions.
Do they, really? Because putting data centers in space would mean multiplying the infrastructure cost by a few orders of magnitude, while being far, far away from cheap energy - photovoltaics would work, certainly, but it will take a lot of it, and it's not like you can just slap panels on the roof - easy cooling, and people.
It's a ridiculous idea, and I don't believe it's what they are really pursuing.
I still have no idea how that would work. Imagine launching an entire data center building into space, and then imagine also launching a solar array to power it, and then also launching a gigantic radiator to cool it... and the radiator is full of some kind of liquid that can never leak even though it's in a vacuum.
Like, sure, but also, that seems like a lot of work, a lot of extra cost, and a lot of risk, all just to avoid building it in Kansas.
It was also hard for many people to imagine a reusable booster, a belly flopping Starship, catching the largest booster ever built with “chopsticks”, a 10,000 satellite constellation, etc.
Orbital compute is technically very feasible. We’re not talking about a datacenter-sized structure, but a lot of rack-sized satellites connected by laser links. SpaceX has gotten pretty good at building, launching, and managing large constellations.
Economically it obviously it has challenges, but there are some advantages (6x solar production, free real estate, less regulation, arguably simpler cooling) to balance the extra costs (launch, radiators, lack of access for maintenance, limited lifespan, etc).
>It was also hard for many people to imagine a reusable booster, a belly flopping Starship, catching the largest booster ever built with “chopsticks”, a 10,000 satellite constellation, etc.
I don't actually think this would be hard to imagine. I've been a huge fan of Space X since it's launch exactly because these types of things do seem feasible because they save so much of money if they are achievable.
A moon base with a secondary launch site, yes. Mining asteroids for precious metals, definitely. I'm not some Luddite.
My only point here is that you can build a data center on the ground trivially easily. Any data center that can exist in space could much more easily exist on the ground... where you can update it and fix things that go wrong. The only issue is politics. I'm entirely happy to be wrong here. If someone can explain the thesis, I'd be happy to get on board.
Solar power is already one of the cheaper (and cleaner!) forms of power generation. In dawn dusk sun synchronous orbits where satellites are always fully illuminated the panels will produce around 6 times as much electricity as those on the ground. And you don't need batteries to operate 24/7 (and as a bonus, the satellites will follow the work day demand curve through the day, reducing latency, at least for some people).
A fully reusable Starship will drastically change the economics of both initial launch as well as maintenance. I expect it will become more feasible to send vehicles to refuel/repair/replace components and keep satellites flying longer. Especially for orbital compute where there will be relatively few dense orbital planes. SpaceX showed modular servers in their video https://youtu.be/k3Un1TizSNg?si=14-bjxXkiyM6cxpg&t=36
First off, let’s not pretend rocket launch dependent solar is “cleaner.” Be reasonable. Solar + saline batteries is pretty damn clean.
Yes, I watched the video. It’s talking about solar and radiators. I agree, if we can solve solar and radiators, it’s feasible tech.
I’m still not entirely sure it’ll be competitive with solar data centers on the ground. I realize I’m no expert, but it just seems like a bizarre way to do computing.
Everything I like about Space X is about doing things that you can't do on earth, because you can't do them on earth feasibly.
> Everything I like about Space X is about doing things that you can't do on earth, because you can't do them on earth feasibly.
I agree those kinds of things are more exciting, but the other angle is SpaceX needs an excuse to do lots of Starship launches, just like Starlink has done for Falcon 9. If orbital compute can at a minimum be profitable enough to pay for the launches it will help bring the cost of Starship down and the reliability up, allowing SpaceX to do more with Starship.
(Of course if orbital compute merely breaks even then maybe the current valuation isn’t justified)
> catching the largest booster ever built with “chopsticks”
I've heard this often. It's not what happens. More correct to say that engineers (not Musk) got a booster to land in a pre-specified spot. The "chopsticks" aren't waving around to catch anything flying by. The booster comes to them.
LMAO, ok, how about “precisely landing the largest flying structure ever built such that it could be grabbed by two mechanical arms attached to the launch pad”?
That will make it clear it’s not actually a giant Mechazilla robot reaching out to grab the booster using literal chopsticks where ever it happens to come down.
You should actually stop and try to imagine it, or failing that, read some of the proposals from companies who want to do it.
None involve launching buildings.
You can look at their models for comparisons with Kansas. There is literature and papers you can read about this. Some go back decades if you include space power transmission, which are related.
That's the cool part, scoofy. You don't need to understand how it will work, you don't need to take any physics classes, round up enough investors, seek out and explain the basic ideas with the people who will make it happen, or invent anything new. Nor do you need to understand political sciences, taxation, jurisdictions, supply chains, or anything else needed to understand the question behind the Data Centers in Space solution.
You aren't even being roped into it with taxes, nor do you have to buy a single share. Other than willingly reading about it on whichever news sources you choose, your observed life will not change a single bit.
You can choose to seek out that info, or you can remain blissfully ignorant. But please don't join the online cacophony of people polluting the threads thinking everyone wants to understand just how ignorant they are.
I get it, I really do. It's a hard task and you don't understand it. But WHY do you feel the need to share that you don't understand? Do you think it makes you look smarter? Do you feel like you fit in more now? If you seek to understand, why aren't you asking questions??
>You aren't even being roped into it with taxes, nor do you have to buy a single share.
Because of the eventual index inclusions, and insane market cap, this affects nearly everyone with a retirement account. Unless you aren't tracking big indexes for some reason.
And some of our compute will run in space, the point is that it's all happening in the background. Most people have no idea how their retirement accounts work beyond knowing it's a bunch of companies pooled together. You don't need to understand how stocks get traded through Alternative Trading Systems, how the companies can decide to take profits vs paying out dividends, etc. A lot of the information is freely available, but you don't need to understand it or take any action.
It's much more likely SpaceX will continue building more ground data centers and using their sat relays to make global connection faster than ground connections can allow.
Yes but the statement is in fact a milestone to meet in order to vest Class B stock options, specifically SPCX needs to put 100 terawatts of compute [1] in outer space and beam it back to somewhere, my guess is Earth.
There's even more rewards for putting a million people on Mars and reaching a market cap of 7.5T by a certain date. Oh yeah he has to stay employed too.
From the SEC Form 3 filed June 12th:
1) This Form 3 does not include 1,302,072,285 shares of restricted Class B Common Stock issued to and held of record by the Reporting Person, which may be voted by the Reporting Person, and the vesting of which is subject to the satisfaction of certain performance and other conditions. 1,000,000,000 shares of restricted Class B Common Stock vest upon (i) the Issuer's achievement of specified market capitalization milestones across 15 equal tranches ranging from $500 billion to $7.5 trillion, with each milestone reflecting $500 billion in additional valuation, and (ii) the Issuer's establishment of a permanent human colony on Mars with at least one million inhabitants, in each case, subject to the Reporting Person's continued employment ("SpaceX CEO Award"). 302,072,285 shares of restricted Class B Common Stock vest upon (i) the Issuer's achievement of specified market capitalization milestones across 12 equal tranches ranging from $1.065 trillion to $6.565 trillion, with each milestone reflecting $500 billion in additional valuation, and (ii) the Issuer's completion of non-Earth-based data centers capable of delivering 100 terawatts of compute per year, in each case, subject to the Reporting Person's continued employment ("AI CEO Award")
The way people just casually use that word again now is so sad. And I don't even mean that in an "I'm offended" way, but more of "I'm embarrassed by the way you're trying to be offensive" way.
> I'm embarrassed by the way you're trying to be offensive
Oooor, try this one on for size:
What if they're not out to cause offense and the malice you impute is just an illusion under which you yourself are laboring alone? What if it was a well understood and not particularly offensive vernacular usage from before people decided they ought to spend their time being offended on behalf of hypothetical listeners?
Why use a word that has some offensive quality to it when other words would be just as effective in communicating whatever you're trying to communicate? You're actively making a decision that you know will cause some level of offense. So the only conclusion I can make is that some level of offense is intended.
In 2004 I used to volunteer as a tutor at an afterschool center in a low income housing project. One day a middle schooler was complaining about how much homework they had and I ribbed them a little, "oh, poor baby."
They were stung. "I'm not poor!" I felt so bad about it that it's stuck with me all these years. Does that mean because I've seen first hand how hurtful it can be that I should chide people whenever they use the P word?
"Chide" is not the word I would use because there is a very obvious difference between the offensiveness of "poor" and "retard", you obviously know that. But yes, if I heard another volunteer at a program for low income kids use "poor" in that offhanded context and I saw the pain it caused in those kids, I think it's reasonable to take that volunteer aside and say "be careful using terminology like 'poor' as it can be surprisingly hurtful to kids that are self-conscious about that". You can do that sort of thing in an informative and compassionate way without being "chiding".
And that analogy isn't even accurate because I'm not the one informing you that the word can be hurtful. You're using it already knowing that. So a better question is did you continue to say "oh, poor baby" to the kids who were hurt by your original comment?
Why would you think I'd continue to say it after realizing I'd inadvertently hurt the kid's feelings? You are making assumptions of ill will from me in the anecdote I shared just like you are making assumptions about the OP intending offense because you didn't like their word choice.
>Why would you think I'd continue to say it after realizing I'd inadvertently hurt the kid's feelings?
I don't think you would continue using it, that was the point I was making and it sounds like you now agree with me that we shouldn't be knowingly offensive.
And my point is that I went out of my way not to use it any more in that circumstance.
Yet in the years since, I still talk about poor decisionmaking, poor luck, poor performance, and poor word choice. Because it would be poor logic to go through life auditing everything I say just in case a middle schooler with a somewhat poor vocabulary might mistake my meaning.
Which brings me right back to "there is a very obvious difference between the offensiveness of 'poor' and 'retard', you obviously know that."
"Poor" has non-offensive uses so you can continue to use it in other ways. You don't need to advocate for the non-offensive uses of a slur. You know regardless of the context, some people will be offended by its use.
Mercy me, a slur? Retarded also has a non-offensive meaning. It just means slowed.
In fact, while I wasn't around at the time I'd wager that "mentally retarded" came into an official usage specifically because it was a clinical, sterile, bloodless, and utterly anodyne descriptive term. Moron, imbecile, and idiot all were once clinical terms. And then people throw them back and forth at each other to call each other stupid, they gained an offensive connotation and new terms were needed.
In 20 years will you find it absurd if people say that "differently abled" is a slur? Will you say "this is nuts, we literally came up with that term to avoid offense?" I will!
I just showed this to a Black person and they said it was regarded, whatever that means.
More seriously. Is "moronic" okay? It's just an ever so slightly more archaiac way to say retarded. The meaning, the negative connotation, the level of offense—it's synonymous and analogous across the board in its time.
Is it okay because "nobody means it like that, they're just using it as a synonym for stupid?" If so, congratulations, you now understand the other side of the argument better than when you started.
Yes, I think there is an intention to cause or risk offense. On the other end, I think there is an intention to be offended and failure to mitigate.
It is a fairly common conflict that arises as a flashpoint in many areas. Different social and legal theories come up with radically different standards.
If someone has a cold, should they not go shopping out of caution for others? If someone is immune compromised, is it their responsibility to take precautions in a store?
oh, no it's exactly as jimmy valmer puts it, there is nothing against mentally disabled people, it is just it's something so stupid that one can't even decide were to start to describe all the points in which is stupid, so stupid doesn't possibly cut it
Bringing back and pinning the word to not derail the discussion of "mental illness", "mental handicap", "slow learner", etc or its use as an offensive?
I think the main issue is that no matter which word/phrase is used, some people will use it as a slur, and changing it so often causes more issues than it solves.
For what it’s worth, I’m not trying to be offensive or edgy when I say that word with friends. “The grass is green and that thing (random topic) is retarded.”
You know that many people are offended by that word and yet you use it anyway when other words would get the exact same message across without the offense. The only reasons to use that specific word are either the desire to cause offense or to revel in the possibility of causing offense.
If someone refers to themselves by a particular slur, that does not grant you any social leniency to call them that too. Consider that exact situation with any other particular slur.
I have not called musk redarded, i have called the idea retarded.
Same as master branch or master/slave communication idiocy. People that get this hung up really have too much free time on their hand, or have too much to gain in discussing language instead of discussing the actual problem.
For example: see how many commenters here are debating words instead of debating the validity of yet another hype-inducing value-pumping statement for clueless investors and fanboys.
On local fintech radio the week it was first announced they spent 3-4 days discussing the financial implication, then just one discussing the feasibility. Guess we're lucky we got one whole day of engineering discussing things for a change
Musk using a slur about himself doesn't grant permission for others to use it. Would monegator have used the n-word if Musk had used it to describe himself? Hopefully not. And let's be honest with each other: Musk says things like this for shock value, because it's not an acceptable word.
Space is an abysmal environment for running compute. It offers no real advantages over doing the same thing on Earth, and it's more expensive, too! Energy is far cheaper and more abundant here than in space. And get ready to figure out things like:
- Heat dissipation
- Radiation shielding
- Either the most complex in-space construction ever undertaken, or the most complex distributed computing problem ever undertaken (no, Starlink satellites aren't good enough, we're orders of magnitude away from replicating the speed and reliability of connections within a single room)
- Zero flexibility, zero repairability, zero upgradability. Either it's working, or you make it burn up in the atmosphere with no in-between. Add on that the rationality of sending mountains of precision-manufactured tech containing many uncommon metals only for them to be completely lost. This makes the pricing even worse, in addition to
- Already high costs for designing, building and launching all that in addition to all the extra weight overhead you're taking in components that don't do computation, when the alternative is building a glorified warehouse in the middle of nowhere.
It just doesn't make any sense. It's a project tied up in hype and created solely so spaceflight can be hastily duct taped to the AI investment hysteria. Ask yourself why no one brought this up before or outside the context of AI, despite the lowering of space launch prices and data centers both existing before any of it.
> why no one brought this up before or outside the context of AI
AI compute is different (slightly higher latency is fine for inference, and there's no issue for training), and there has never been so much backlash against data centers or other infrastructure buildout. If an increasingly-non-minority of politicians get their way it will in fact be cheaper and faster to get some servers shipped to space than it will be to get the permits and build it out here on earth.
Also, most Datacenter maintenance is just dealing with the problems you see with space- power and cooling. Solar is 5x better in space and a lot more consistent. Here on land they're shutting down nuclear and imposing so many new regulations on gas and coal (and now on solar and wind) that there aren't many grids that can support growth at the scale that's being requested.
Nobody is claiming that all compute is going to space, which is what you seem to think you're arguing against. There's high-dollar demand for it right now, so if we want to be multi-planetary it's the perfect time to start tackling the "compute in space" problem that needs to be solved. Or you need to prove to future people that "All compute can be forever located on Earth" which seems a lot harder sell.
But it'll have to start with Sinophobia so nobody wants to touch it. LEO objects are just such a non-issue that you'd have to really be grasping at straws (though the "Data centers will consumes earth's water" people can probably fall for it)
Assume reusable spaceflight eventually brings launch cost close to the cost of fuel. This is close to happening.
The overhead of building out grid and power infrastructure on land would then exceed the installation speed and cost relative to space based deployments.
Also assume the compute that does make it to space has a short shelf life anyways so lack of ability to repair is a non issue. As we scale manufacturing on land this will increasingly be the case.
China has already run experiments and served models from space, so we know the heat dissipation equation is solvable.
Finally you’d arrive at a similar model that’s already proven successful with Starlink but applied to serving inference.
The key question is speed to scale new deployments to meet demand. If the markets demand is near infinite, they will choose to fund space based deployments over slower land deployments.
That makes no sense to me. You've cited one overhead for building on land but just didn't bring up the billions of overheads for deploying satellites. The associated costs of launching this stuff into space blows the price of boring old power grid infrastructure out of the water, even with launch prices assumed to be the cheapest they can be. For instance, the extreme additional costs of equipping each payload with their individual systems of power generation, flight computers, maneuvering systems, structural components, communications infrastructure, heat dissipation hardware, radiation shielding and so much more. You have to pay the price in weight and money for all of these inefficient small-scale components for each unit launch, compared to plugging in an additional server rack and taking advantage of the economies of scale in a centralized data center that already solve all of these things. And that's before you get into the costs of designing, mass-producing and operating these things, compared to the costs of running a normal data center. And that thing about reusability that I mentioned also cuts into the margins. Data centers can keep working even if their hardware is a generation out of date, they can sell off old equipment, they can repair individual components instead of throwing out a whole rack. This is literally setting your GPUs on fire. It's just one more thing that makes it even more expensive.
Heat dissipation in space is possible, of course, but every kilogram you spend on heat is a kilogram you could've spent on something else. When you're talking about boxes that generate so much heat, you're going to need to spend a lot on that ancillary hardware in each unit that, again, makes it even less rational.
Then the concerns about megastructures or distributed computing go unanswered - to my knowledge, we simply don't have the technology for either of these right now. Starlink isn't close to solving it - the bandwidth of a Starlink satellite is nothing in comparison to the bandwidth of a single current-gen server GPU connection.
Energy is not cheaper on earth. Solar in space gets 24/7 power at a 30%+ rate vs the surface. Radiative cooling is passive and cheaper because there are no HVACs or chillers and high temperature chips reduce the need. The ISS does this already. Radiation really isn't a big issue with ECC and redundancy and optical links are fine for batch training.
The only hard part making the math work is the launch cost. They need reusable and reliable starship economics. If they hit that goal, it will become cheaper for pre-training, which is 70%+ of the budget for the frontier models.
> It offers no real advantages over doing the same thing on Earth
Abundant solar energy, free real estate, less regulation, less backlash from NIMBYs, simpler (yes) cooling.
> Energy is far cheaper and more abundant here than in space.
Huh? The sun is obviously the most abundant source of energy in the solar system.
Satellites in a dusk dawn sun synchronous orbit can be fully illuminated 24/7, so they receive ~6x more solar energy than panels on earth. They also don’t need batteries to operate 24/7, and the panels don’t need glass to deflect hail.
> Heat dissipation
Yes, it will require large radiators. They’re mechanically simpler than terrestrial cooling though.
> Radiation shielding
It turns out generative AI is somewhat uniquely robust to occasional bit flips.
- Either the most complex in-space construction ever undertaken, or the most complex distributed computing problem ever undertaken (no, Starlink satellites aren't good enough, we're orders of magnitude away from replicating the speed and reliability of connections within a single room)
It won’t be a single structure, and will only be used for inference, so latency between satellites/racks doesn’t matter.
- Zero flexibility, zero repairability, zero upgradability. Either it's working, or you make it burn up in the atmosphere with no in-between. Add on that the rationality of sending mountains of precision-manufactured tech containing many uncommon metals only for them to be completely lost. This makes the pricing even worse, in addition to
- Already high costs for designing, building and launching all that in addition to all the extra weight overhead you're taking in components that don't do computation, when the alternative is building a glorified warehouse in the middle of nowhere.
I think people vastly underestimate how much a fully reusable Starship will change the economics of space operations.
Not only the initial launch costs, but things like refueling and repairing satellites becomes more economical. I wouldn’t be surprise if SpaceX sends Starships to refuel/maintain satellites to keep them in orbit longer. In fact, SpaceX’s own animation shows modular servers sliding in and out of the satellites.
Space-grade photovoltaics are >10x more expensive than ground based panels. Add some (Tesla) utility scale batteries and it can run 24/7. No need for expensive radiators or rocket launches. And personnel can upgrade the hardware every time there's a new generation of GPU's.
Putting datacenters in deserts around the equator is a much better idea than in Space. If you're really optimizing for cost that is. If you're optimizing for SpaceX meme-stock valuation the former wins
A global centralized internet provider? I mean, just that, might be a trillion dollar company. Let alone build out datacenters? Those are not easy or cheap here on earth. You can build out datacenters at scale with minimal need for power or cooling. I think the current rate for a data center (not gpu) is 120million. Datacenters are super hip now too.
Elon has always been a Department of Defense/intelligence asset larping as a capitalists. You can't exactly go to congress and get a quarter of the government budget to do globe spanning surveillance from space, or convince them to let you fund, build and maintain a platform for psyopping your own civilian population (x) So you go get your illegal projects funded through capital markets with through an asset like Elon. As soon as the department of defense started focusing on subterranean warfare and how its the future of wars Elon starts a tunneling company.
Elon is just Howard Hughes v2, they're running the same script again.
Is the TAM for airlines finite? They ship an indefinitely growing volume of cargo and people. Reusable rockets will be no different. Cargo into space, people in point-to-point orbital flights and to Mars.
Not a sound argument because you would have said the same thing before they did Starlink. "The TAM for launch services is finite. There are only so many countries/companies that want to launch satellites."
Musk would argue infinite. They literally want to create offworld colonies, with everything that entails. Obviously it's crazy, but it beats the pants off more adtech.
I'm bullish on DC in space with laser links. The whole sentient sun/railgun on the moon... hey, go big or go home. I would have probably just asked MBS for money on that one, and renamed the railgun "the line (of ketamine)".
It’s insane enough to consider undertaking the risk of a moon or mars colony. But doing so with him as the sole load bearing link in your supply chain? What if mars turns out to be woke and you’re fed into a wood chipper like USAID?
Last week a 13 year old video of ceo of Ariane Airspace got popular on twitter. When asked about spacex and reusable rockets he said: "there are only 25 satellites launched a year, every year, and that’s not going to change"
Currently a single Starlink launch is 25 satellites. And there are 100 such launches a year.
If only the shareholders had any kind of voice - if only it was illegal to issue a fake IPO where you sell an overwhelming number of shares stripped of their voting power - if only the market responded rationally to this boondoggle.
If we're going to right the ship in turn of common sense a bunch of people need to lose a bunch of money, I just hope it doesn't mostly hit passive investors and instead lands mostly on Elon-stans.
If only you there were companies who could choose which type of shares they'd like to offer...and if only people could buy the types they want from the company that aligns with their ideas...
Passive "investors" can go and invest in ETF or whatever else that does not include company shares without voting rights.
How is it legal to have different share classes? You could make 100 shares that can vote and then sell 99.9% of the company while maintaining full control. Seems strongly against the spirit of a publicly traded company
I mean that just begs the question how we got to this point. And how far detached is the public market from real value. Many shares don’t confer actual value. Hell a lot of these companies don’t pay dividends in addition to very little or no voting rights.
Dividends are an interesting topic. There are sort of two ways to view the stock market - one is as a loan with a very drawn out repayment schedule that may be amicably adjusted for the health of the company (that one relies on two repayment strategies, dividends, buybacks and acquisitions). Once a company has issued stock the only two real reasons to care about the value is for the utility as a valuation tool which has steeply fallen off with indepth audits becoming the norm and, of course, further stock issuance - a healthy share price allows less dilution for further issuance. The other view of the stock market is just pure gambling - if you take away the disbursements and shareholder voting control then all that's left is the valuation which we can do better in different ways (and stock price is often disregarded when considering loans and M&A except when the difference can be arbitraged).
Really, it's kind of all meaningless BS, but as long as most people believe it isn't then it's a great way to grow assets by doing nothing which is an activity that we apparently want to encourage?
I wouldn’t see it as just pure gambling. I’m not a dividend queen that really cares all
That much about it. I do think we are sliding along the scale from shares having intrinsic value as an ownership stake in an income producing asset to collector cards that we track societies valuation of a company with. And at some point in the next 50 years shit will hit the fan. Unfortunately sitting it out isn’t really an option either
There's some companies (Snap being one of them) where certain classes of shares have NO voting rights. If you think SpaceX is lopsided, look at Snap's shares. They have A, B, and C class. The founders of Snap own 95% of the voting through their Class C shares... literally only pre-IPO folks have any voting rights.
Musk didn’t invent the idea of using multiple share classes to ensure the founder(s) retain control of the company, see Rupert Murdoch, Google, Facebook, etc
From the regulators’ perspective: it is a risk, but you disclosed that risk in the prospectus that buyers are assumed to have read (what percentage ever actually do?), hence it is fine
Well, when you buy into an IPO, they make you sign to say you read it. So either you did, or you made a false statement on a legal document
I suppose they kind of do, they could sell the stock and drive the price down. That wouldn't force Musk to change direction, but it would hurt his wealth.
If you have this much wealth in something as unpredictable as the stock market I think the very wise move is to start converting it into stock backed loans[1] that you leverage to buy real assets - buy islands and small countries, buy other successful large companies and just let them keep trending slowly upwards, pick up that New Zealand bunker... just try and get your money diversified and into things that would survive a stock price collapse. Even if Elon did this with 1% or 10% of his wealth and kept the vast majority of his assets invested in this one clump of companies and they went to zero he'd still be sitting on an insane pile of cash.
1. Lets all pour out a drink to the bankers who authorize such insane loan agreements.
With its IPO, SpaceX secured its role as the vehicle for consolidating Musk's vanity businesses into one closely held public organization that can more easily convert publicity into investment and internally reallocate funds and debts based on his personal whims.
So yes, SpaceX is pivoting, but it's to no one's surprise.
Buying Tesla is just buying SpaceX. When they roll up Tesla, it will be in a stock deal not real money. None of this is real money and I would argue that Musk is not the world's first trillionaire because there is no reality in which he could get that money out of SpaceX
I mean, if he wanted to sell tomorrow, who COULD spend $2-3 Trillion to buy it, and who WOULD? Anyone with that kind of money to spend today knows what a scam it is
Sure, it's reasonably likely they merge at some point but the ratio between the prices at that time will be different than the current ratio, so it still matters which share you buy today.
Especially since Musk is heavily incentivised to merge them at a price favourable to spacex since he has a much larger share
Musk may not have a few billions liquid on his cashing account, but he have something better: collateral to raise a few billions for everything he may want.
I mean as far as I know no human being is even a billionaire anyway if you only count cash. It's one of the things the "eat the rich" crowd is particularly bad at internalizing, that there's a difference between value and money that can be spent on food or hospitals.
It is much more feasible for Jeff Bezos to sell a billion dollars worth of Amazon tomorrow or Bill Gates to sell a billion dollars worth of MSFT tomorrow than it is for Elon to sell a trillion dollars of SPCX even over a year's time
I get that net worth is more than just cash, and that is not what I meant and it's pretty obvious that isn't what I meant. It may not just be cash on hand but if an asset is completely illiquid at it's purported value, is it actually worth that?
It is, because of the time component of money. Money is a way of storing the value of labor in a manner less affected by time. If you sell it all at once, its value is greatly reduced. All investments include that time component. Point-in-time net worth doesn't have much value as a measurement in part because of this.
Somewhat doubtful. (The first part of your statement, anyway. Of course the difference between abstract "value" and hard, spendable "money" is a thing.)
Like Mr. Hanson said in my sibling comment, some rulers are (or were!) bound to have amassed incredible amounts of resources. For historical/non-present-day examples, consider looking into figures like Jakob Fugger or Mansa Musa.
I have no idea why the adjusted net worth of people who died centuries ago would have anything to do with money in the modern day. And at a glance, neither of your examples had billions of dollars in cash, which is the point I was making. They may have had a lot of value tied up in illiquid investments, which is exactly like Musk's valuation.
Cursor's edit data is invaluable to anyone who wants to train a coding model. Probably the best data available outside Anthropic and OpenAI. Coding models are seen by the leaders in AI as both the biggest current revenue opportunity and the best way to accelerate the progress of AI and bring about recursive self-improvement that will create superintelligence. So yeah, it's easy to see how SpaceX could see it as in their interest to purchase Cursor with 2% of their equity.
Mars was never going to happen without revenue. Starlink is providing revenue but probably not enough to build a whole city on Mars within our lifetimes. SpaceX needs more and AI is the only near-term way.
> Cursor's edit data is invaluable to anyone who wants to train a coding model.
Ok. So what prevents a company from offering a Claude Code/ Cursor equivalent, with 100% subsidised Claude (= 100% free), capturing the exact same data that Cursor does? If the data is worth in the tens of billions, the cost of subsidising the usage is negligible.
Cursor also has a large customer base including most of the Fortune 500, talent, and their own coding model and training infrastructure using their data. You wouldn't get those automatically by subsidizing Claude, and the many months or even years it would take to ramp up your own IDE/coding hardness from scratch and acquire customers to match Cursor would put you way behind in a field where the SOTA climbs every month. It's also not clear that Anthropic would let you undercut everyone else using their API; they could cut you off at any time. That would be a very precarious position.
The video explains that it is spelled out in the prospectus that SpaceX is counting 70%-80% of their total addressable market to be AI related and only about 7%-8% to be space-related.
I saw that too, and it's so depressing. SpaceX was pushing the envelope of "interplanetary" travel/species, and to see it being reduced to a 7-8% "side-quest" :-(
> was pushing the envelope of "interplanetary" travel/species
I don't think they were. All the Mars stuff is just dressed up version of The Boring Company - a distraction by Elon to better position his other interests.
There's no money in going to Mars and there's no reason to, from a financial perspective, and Elon doesn't care about anything beyond wealth and power.
Safari isn't even open source, and Chrome has never been a fork of Safari.
But yes Blink definitely started as a Webkit fork, and everyone would have found that laughable if someone bought that a proprietary fork of Webkit for $60B.
Cursor's Composer 2.5 is one of the few models out there focusing on coding, which is the one thing most of us here want. It's pretty good! It's not near frontier level insight generating genius, but it's regarded as very capable and trustable, and is indeed a lot better than previous Kimi. It'll be interesting to compare it versus Kimi 2.7 Code, which just dropped, which is also notably a coding specifical model. I'm expecting we'll see more of this over time and I think it has huge rewards, and Composer 2.5 is early proof.
I'm not super concerned about the spend to train the model, especially given that Kimi was famously incredibly cheaply made, and given what they are competing with. I don't think that's a meaningful concern.
Reciprocally, and in far more important relevant in my humble opinion: in terms of cost to run models: Composer 2.5 is easily one of the cheapest models out there. It's fantastically cheap. It's token efficiency is through the roof astronomical. I think this training for a coding specific model has yielded something incredibly special here, and I hope SpaceXLAIC isn't the only company doing this.
However it definitly isn't _just_ Kimi. The weight will be different after that 85% of extra training on top of the base model.
If those different weights are better are worse doesn't change that it's in most meaningful ways not the same as the base one.
I would encourage you to lookup their blog posts about their post training process if you want a bit more faith that they aren't running an extra 85% of compute and burning money with no-ops.
I don't think it's all no-ops. Still don't think it's a particularly relevant model/company/product.
I'll defer the reading until I see signal that they have something worthwhile. I've watched a couple interviews and used the product, neither of which impressed me.
>One of the things that makes @SpaceX
so valuable is how valuable it is. The Cursor acquisition costs materially less in dilution because of SpaceX’s high valuation.
They are buying it with overvalued stocks, so it isn't real money. Probably the Cursor team will be able to sell it when the SpaceX stocks will be already crashed.
Why are you comparing it to building expensive modern hospitals? Why don't you compare every other tech acquisition to that? because that's not a relevant comparison, and building expensive modern hospitals has nothing to do with the goal of for-profit corporations.
SpaceX has 3 major businesses: Space, Starlink, and AI.
This is the same straw man argument of "country Z has N homeless/unhoused people, why are they building a space/military/education widget for $Y price, when they haven't found homes for N unhoused people yet?"
It's not like they could easily cash out all of those $60B. I always find it troublesome that we generally conflate cash with stonks, market caps, and such.
> It's not like they could easily cash out all of those $60B.
They probably have a vesting period of some sort (as they would with cash as well) but beyond that they will definitely be able to cash out all of their money as soon as they are allowed to.
$60B is 3% of SpaceX at today's valuation, Musk had no issue selling this amount of Tesla shares to buy Twitter. The idea that stocks are somehow not liquid is an nonsensical urban legend.
Elon is consolidating all of his property into one single megacorporation because he is confident that nobody will ever challenge this, given the current political direction of the United States.
> the current political direction of the United States
These kinds of comments reek echo-chamber parroting and zero substantive research. As someone that very much enjoys and carefully follows politics, the current political direction points squarely to Republicans getting absolutely pummelled in the midterms, effectively turning Trump's administration into a 2-year lame duck. What are you even talking about?
Obviously you're trying to be snarky, but I hope you realize that Congress does, in fact, have (fairly broad) statutory authority over executive agencies[1].
Can it change? Yes. That's not the same question as will it change. And that is also not the same question as "will the change result in a different posture towards antitrust."
When was the last time substantive antitrust action was taken that forcefully restructured a large company to a significant degree?
And I hope you realize that while Congress has authority over a lot of things, that authority is being routinely overridden by the current presidential administration, including fundamental things like spending and declaring war.
So it comes across as a bit foolish to assume that any Congressional authority actually exists, or will continue to exist into the future, since we have many examples now of where that authority seemingly doesn't matter anymore.
Especially since the majority of Congress is in the same party as the current President, and is making no effort not to cede congressional authority to the executive branch.
Congress isn't being overridden at all. Congress has the power and authority to rein in Trump whenever they want. They are choosing not to, because congress is controlled by the Republican party, and the Republican party is currently ducktaped to Donald Trump himself, and even barely delaying anything Trump wants gets you primaried.
But mostly they are fine with what's happening so they have no desire to stop him.
One of the things I find really interesting as a Brit is hard to put into words:
Americans tell us that we don't have a constitution — when we do, it is just not wholly written down. (It is in part).
We have a constitution that is flexible and precedent-based, but pretty stable, and it has emerged on top of the bits that are written down, and has amended them over time (for example, it is built in part on Magna Carta, but only two or three of its principles remain in law.) Notably a bit more of it got written when we agreed to be bound by the ECHR, but that was mostly absorbed into our understanding.
It has taken us hundreds of years to get to this stability, and it is defended from attack from pretty much all sides; every government risks changing it and there is pushback each time, because you can't govern if there aren't rules. The rules are precedent and convention, and there are various authorities and archives that are consulted to work out what they are if people think they are at risk.
We are regularly told by Americans that this is an intolerable thing; we need a written constitution or we can't know what our rights are!
But those same Americans, right now, are engaged in exactly this process. You have a set of written rules that give Congress power over things, and you are currently evolving a set of precedents that suggest that the executive can simply wander past them and Congress somehow shows deference or refuses to assert its power in some situations.
You're right at the start of building an uncodified constitution on top of the old one just as we did on top of Magna Carta.
It's not entirely new to Trump; every President in my lifetime has pushed on this except maybe Carter. And sometimes they push back (Roe v. Wade was part of this uncodified constitution and probably needed to be a written amendment.)
It could work out but it's important to understand that is what you're doing. And it's not just Unitary Executive theory or presidential immunity; the emergence of the Supreme Court's "shadow docket" is emblematic of the same process.
We had Boris Johnson and Liz Truss; the fact they happened at all demonstrates our system still has vulnerabilities.
I’m no fan of venerating a written constitution either, but I have an anxiety about parliamentary-sovereignty: events like Brexit and BoJo demonstrate a need for something to bind parliament, lest we get a 21st century Oswald Mosley. Right now we have… decorum and monarchy: a firewall made of damp serviettes.
Prior to Brexit, I feel there was some (if vague) kind of accountability from parliament to the EU - which is why/how Brexit revealed the fault-line in the Tories that split the proto-fash types from the amoral-business types to give us Reform/Restore today, and they’re uncomfortably popular. So what I’m saying is that Brexit weakened our accountability mechanisms more than people realize as the national-conversation focused only on economic impact and immigration.
In summary: do not convince yourself that “it cannot happen here”; nor that current safeguards (if any) are sufficient.
Oh I am not convincing myself at all. Like I say, every government ultimately ends up risking it and getting yanked back by the Commons (or the Lords in the case of e.g. 42 day detention of terror suspects, various other bad overconfident schemes like ID cards).
It's a pattern, a cycle. It is always under a sort of measured level of threat. But it is largely stable because at any time the party in power always has an iconoclast (Heseltine, Cook, Davies etc.) who is willing to get in the way, and there is essentially a Commons (or Lords) tradition of making sure that person gets heard.
And because the Prime Minister isn't really that special — just the loudest voice in his party, subject to its back-stabbing — there is a pattern to their failure, too.
(There is a bit of a concern that we're getting too used to them not lasting more than a couple of years, but I do get rather frustrated at the both-sidesing that is including Starmer with a run of utter inadequates from the other side)
Again the point I was making was only that the USA is getting a little further down the line towards wrapping up the clarity of the written Constitution with a layer of yes-but convention... It gets a lot more difficult from here.
There is a narrow window to pull it back, but as much as I think they are for the rule of law, I don't think the Democrats, even with potential rule-of-law-asshole injections like George Conway, are truly prepared to hew that close to the Constitution themselves.
Even if Congress may reverse things in the future, there are many opportunistic things happening right now, and it seems like the spacex situation is one of them. The emerging picture feels like a 'flood the zone' strategy (not by coordination, but by practical effect).
While other commenters have pointed out lots of details that point towards the favorable structural environment going forward, another idea that roots my thoughts towards this is that by creating facts on the ground, they are defining the new starting point.
Ultimately, reversing all of the different 'wrongs' or irregularities will be costly to both the opposition's political and attentional capital.
> the current political direction points squarely to Republicans getting absolutely pummelled in the midterms
Even if so, are the Democrats really going to do the house cleaning required to fix this? Their recent history implies that they'll try to pretend things are running normally, until it all explodes in their face (again). Maybe I'm wrong, and they'll actually fight for the country, but... I'm not surprised that companies (and markets) are expecting them to just... not.
Brother I'm not clear there will be a midterm election and I'm confident most of the electorate on both sides feel very disenfranchised voting and suspect the system is not working on the level.
Going to do my best to respond to this while still following the HN guidelines:
> Please don't use Hacker News for political or ideological battle. It tramples curiosity.
RIMR says:
> nobody will ever challenge this, given the current political direction of the United States
It's obviously hyperbole to say that NOBODY will EVER challenge this, but I'd say it's directionally correct:
1. The Supreme Court is controlled by a conservative, pro-big-business majority that makes it very difficult for any legal attempts to challenge Elon's actions to survive litigation.
2. The United States Senate has a conservative, pro-big-business bias due its over-representation of rural voters and its internal norms (filibuster)
3. The United States House has a conservative, pro-big-business bias due to the gerrymandering efforts of Republican-controlled state legislatures across the country (which the Democrats have tried to counter and failed, see Virginia)
4. The conservative, pro-big-business Supreme Court has ensured that elections in the United States overall have a conservative, pro-big-business bias due to the unfettered spending allowed after Citizens United.
So yes, the winds seems to be against Republicans and Trump in the mid-terms, but the structural biases of the government are still very much pro-big-business, pro-capital, and anti-regulation.
It will take much more than a single mid-term cycle to reverse that trend.
> the current political direction points squarely to Republicans getting absolutely pummelled in the midterms,
I am not this optimistic and I think it is dangerous to be this optimistic. Even putting aside mischief, just as a matter of reality, the gap always closes. A loss, yes. Hopefully. A slap around the face, maybe. Perhaps enough to get a clear win in the House. A pummelling? Nah. I think they might still keep the Senate.
Trump supporters falling out of love for him could well just lead them to focus their attention on down-ticket Republicans who figure out how to make them feel OK about their past choices, and since they need to feel OK, they will come out. Wholesale rejection of the party is unlikely since it is already splitting into two factions.
is spacex a space company? I thought they were an internet provider that wants to use their strategic advantages to get into AI including AI infra like data centers.
SpaceX since the IPO is an AI company with two side projects: social networks and space.
I say this based on their filing which says that the vast majority of predicted profits will come from their AI company, citing a $36.5T total addressable market.
SpaceX isn't a space company anymore, it's an AI company. In their IPO filing, of their projected $28T total addressable market, only $370B (~1-2%) of that is from space [1]. The rest is primarily AI, with a sprinkling of Telecom revenue from Starlink.
Given xAI's Grok is way behind ChatGPT & Claude on coding capabilities, whereas Cursor was able to get in spitting distance of them w/ Composer 2.5 by simply running post-training on Kimi K2.5, I'm not sure Elon could dream up a more perfect strategic fit.
Cursor likely has the largest, highest quality dataset of any private firm for training new coding models, which would compete SpaceX's trifecta of becoming a viable competitor in the AI race:
1. Access to compute (they have so much that they're renting capacity to Anthropic & Google)
2. Liquidity for R&D+M&A (largest IPO in history)
3. High quality training data (this Cursor acquisition)
> Isn't it kinda bizarre that Elon is pivoting SpaceX to something else?
In a vacuum, absolutely yes. But in the bizarre context of the AI economics, chaotically scrambling to bring everything you need to compete in-house makes perfect sense.
Arguably, when compared with either OpenAI, Anthropic, or Google, SpaceX/xAI now own the most compute, are the most financially liquid, and (assuming the Cursor acquisition goes through) have the largest corpus of high quality training data.
We may very well be a couple of months away from a Grok release that goes toe-to-toe w/ other American frontier models, IMHO.
So when you look at this as a $60B play to capture an additional 10-20% of an estimated $26T total addressable market, it makes a lot more sense. Now, whether that projected TAM is even remotely close to reality (or even just enough to make Cursor worth $60B) is another question entirely.
No, it really still makes no sense. Where's the moat around what Cursor provides? If Cursor is really that great, surely something equally great could be developed for a measly $1 billion or so? Is it brand recognition? An established customer base? Surely they don't have $60 billion worth of either.
The key point I think you're missing is Cursor's "moat" isn't around their product or brand, it's around the gigantic corpus of usage data they've almost certainly collected.
It is simply not feasible train an LLM to be as good as these frontier models are without a TON of high-quality examples of what "good" looks like. Every time a Cursor user (who didn't opt out of analytics) does/doesn't hit a "retry" button, or rejects/accepts an LLMs output, it allows Cursor to log record of a specific LLMs output and a binary signal of that output's quality.
Given they've been at this since 2022, and for most of that time sat comfortably at #1 in market share among comparable AI coding tools (only recently getting topped by Claude code), Cursor likely has the largest, highest-quality, SWE-specific dataset in the industry by a sizable margin.
Grok being so late to the party could only train on twitter data in combination with whatever they could source publicly or purchase privately, and likely hasn't had anywhere near the usage they'd need to build up their own competitive dataset from scratch anytime soon.
If you believe (as SpaceX seems to) that the AI's total addressable market is over $26T, and acquiring proprietary, high-quality training data is the difference between capturing ~1-2% of that market and ~10-20%, then $60B starts to look like a bargain.
>they have so much that they're renting capacity to Anthropic & Google
they could not make use of it! That's why they're renting, they don't have access to good distributed training software, so the heterogenous cluster they made was bad for training and they don't have enough demand for inference.
>Arguably, when compared with either OpenAI, Anthropic, or Google, SpaceX/xAI now own the most compute,
lol wut. You think they have more compute than Google?
> are the most financially liquid,
You think they have the most cashflow?
>the largest corpus of high quality training data.
What? You think they have higher quality training data than Anthropic, Google, and OpenAI?
>We may very well be a couple of months away from a Grok release that goes toe-to-toe w/ other American frontier models, IMHO.
Training on what? They rented both of their data centers away!!!
Grok 4.3 was made on Colossus 1? It's not frontier quality, but it's definitely not nothing. Also, Cursor brings more expertise to help them improve their utilization further.
Additionally, I struggle to see how renting out excess capacity is anything but good. It brings in a ton of cash at a price premium, and ensures they have headroom to gradually phase out rental capacity as their internal demand increases.
> You think they have more compute than Google?
Fair point. No way they have more raw compute, but I do think it'd be fair to say they have more "excess compute capacity" than Google, but even that is pretty speculative.
> You think they have the most cashflow?
Again fair point. I was overestimating how much cash comes from a $2T IPO before actually looking at the numbers. My revised take is that SpaceX/xAI are now in-line with the other labs on cash liquidity, rather than leading (where pre-IPO they were way behind)
> You think they have higher quality training data than Anthropic, Google, and OpenAI?
Yes -- I'm standing by this one. Cursor has a multi-year head start on agentic coding data collection, and a GUI UX that likely provides richer user sentiment & quality signals than something like Claude Code.
Obviously OpenAI & Anthropic have far larger proprietary datasets for chat histories, and Google is an undisputed leader in data hoarding. But when it comes to agentic coding specifically (AI's most compelling use-case, IMO), I think Cursor's data is a HUGE deal. This is backed up by how good Composer 2.5 was given it was essentially Kimi K2.5 + Cursor data.
Additionally, I also suspect its possible to mine Twitter for user engagement & sentiment analysis to create surprisingly useful datasets.
> Training on what? They rented both of their data centers away!!!
I thought they only rented Colossus 1, but you're right they're also renting out a portion of Colossus 2. That said, they still have 2/3 of their Colossus 2 facility available for internal, which is likely enough to build something competitive. They also have 90-day termination agreements in place once they forecast a need for more internal capacity.
All that said, while I appreciate your pushback & I was overly hyperbolic on a few key points in my last comment, I stand by my core theses: The Cursor acquisition (assuming it goes through) combined with the SpaceX IPO puts xAI/Grok right in-line with the other big labs, at least in terms of positioning.
Whether they're able to execute remains to be seen, but I would not at all surprised to see a frontier-quality Grok 5 or Composer 3 release before the end of the year.
Since Cursor is kind of a Select-Your-Model front end for various AI coding systems, how much of it could be to just make sure Grok gets to the top or default choice of that Select-Your-Model list?
Because cursor gets some of the highest quality training data from the world's programmers and responses from the full ecosystem of model vendors and access to active code bases. XAI wants the data.
Perhaps that's where the money and strategy is. (a) stronger need; (b) if you can build systems without real expertise, you don't have to stomach their salaries or politics.
Umm. It’s real development work in real settings with real model output. That is a high quality dataset. The fact that it isn’t good code from elite engineers is confusing what good means in the context of coding agents. First is how to respond to a range of prompts. For that you need diverse real world conversations. Second is the ability to respond with good code. That is about labeling or other data curation after the fact or other training methods. So it’s a downstream consideration
How is buying a company sending messages on the Internet for $45b in the owner's interest when Westinghouse, who built nuclear reactors sold for under $10b? The market is irrational.
I think it has something to do with the fact that the first company is better at spreading messages on the Internet than the other at building reactors.
It might not be irrational if there is more revenue in people who are willing to spend money on internet messages than there are on those willing to spend for nuclear reactors.
So far, I see way more companies paying each other in a circular fashion than individuals forking out anything near whatever tam SpaceX is projecting for AI.
when you see the list of major investors in Cursor that will never break even if it stays independent, and compare it to the investors in SpaceX, it all makes sense.
Hardly the fault of anyone is it for not reading a 100+page document meant for investors when it literally in the name .
If they don’t want us to think of them as a Space company they could have taken the xAI name (like how grammarly did with superhuman) or called it Musk Inc or whatever else.
I mean, the answer is obvious if you do not deliberately try to put a message in the worst light possible:
- "Space company" has a major LLM+datacenter business called X.ai.
- LLM for coding is a big business, as you can see from trillion dollar valuations of Anthropic.
- Cursor is popular and gives you a headstart on the business.
- Instagram was bought for the price of many many hospitals. Uber is more valuable than companies owning the cars. Different business models, entirely different valuation models. Not sure what that comparison entails. You know it. I know it.
Whether it is a good purchase or not, we may not know, but we know your characterization is just outright dismissal without much rationale behind it.
I bought a small amount (4 shares) of SpaceX stock on IPO day for $160/share. This Cursor purchase does not upset me as a very minor shareholder. Elon Musk seems to have a unique ability to help grow companies that lead to shareholder value. In particular, it seems that xAI overbuilt data centers because their model fell behind. SpaceX could lease data center capacity to Google or other big players (which they're doing), or they can use it internally. Buying Cursor lets them do it internally.
My employer recently switched us all from Cursor to Claude Code. Aside from my personal preference for having a chat window inside VSCode, Claude Code is painfully slow compared to Cursor for my workloads. I think part of this is due to Claude's massive bump in popularity without a similarly rapid build-out of compute. So, the low-hanging fruit for Cursor is to have a massive speed advantage over Claude Code and regain popularity that way. (My current paid AI subscriptions are ChatGPT, Gemini and Cursor. I do not personally pay for Claude.)
And as far as the pivot goes, there seems to be speculation that Elon Musk wishes to roll up all his companies into one big company. So, it doesn't really matter if the AI company lives inside SpaceX or Tesla, since it'll all be one big thing in the future.
Part of Elon Musks strategy seems to sell some kind of hype that does not materialise or at least won't for long (Mars, autonomous Cars) The vast amounts of money collected are then used to develop products that are still a significant progress in its market. Now AI is where all the hype is. It's difficult to sell some hype without AI currently.
> bizarre that Elon is pivoting SpaceX to something else
SPACx designs, manufactures, and launches the world's most advanced ponzis on pyramids. The company will be re-founded every year by Xlon Tusk to revolutionize capitalism, with the ultimate goal of making market multilevel.
spacX has gained VC attention for a series of web3 milestones: It is the only AI company ever to run a 1000B-A3000-Thinkkking on low-Cost toasters, which it first accomplished in May 1945. Grokipedia made history again when its ClosedAI attached to Moonshot Kimi, exchanged token payloads, and returned Alignment to money — a deeply challenging feat previously accomplished only by Cursor 60B. Since then it has distilled cargo to and from Moonshot multiple times, providing regular RL missions for Goog.
SpaceX is just a vibe company. Nothing they are doing makes sense on a valuation basis, which their investors are eventually going to painfully figure out.
I’m worried all this crap will distract and damage the things they are great at (rockets) similar to what happened to Tesla where outside the stock price it’s pretty dismal
SpaceX’s interest is being an Enterprise AI corporation, they identified it as a $24 trillion addressable market, in comparison to their quarter trillion rocket related one
It’s all disclosed in the S-1, you read it right?
In America all you have to do is tell potential investors what you’re doing, its up to the people to use their discretion afterwards
Back in the early days of Heroku (when I worked there), we were all fairly deep into the Ruby community. Ruby has never had a great reputation for performance, but... it seemed like almost a running joke that any time you went down a rabbit-hole trying to understand some weird performance issue you'd eventually discover that @tmm1 had already identified the same issue months earlier, patched it in core, and given an hour long talk about it somewhere. Despite his ability and willingness talk publicly about quite deep technical topics Aman always came across as an incredibly quiet and humble in person. Every Ruby developer has benefited from his attention to finding and fixing performance issues. I'm sure the same can probably said for every GitHub user (where he worked for years).
Congrats to the entire Cursor team! I don't know all of their stories, but I do like to smile and celebrate a little when I see people who are often hidden in the shadows quietly making things x% better for all of millions of us every day for many years getting reward for that effort.
Especially with Elon? It happened one time (twitter) and if you think twitter was appropriately staffed at 8k employees you need your head read. Cursor operates with a very small team, and practically all have equity, so a $60B exit means congrats are in order.
Wow. I wonder (genuinely) how such a mammoth payout will affect retention? I guess the option vesting will serve the purpose?
I guess if I found my already-vested shares were worth $4 million, I'd still be motivated to get the remaining $8M by working a couple more years. How surreal that would be, to find yourself extremely "FIRE" (Financially Independent, Retire Early) but also kind of obligated to yourself to keep working hard for a few more years, as walking away from that kind of money seems crazy.
Usually the grants are restructured. Ie instead of what was vested and unvested rank-and-file gets new 4 year grants. And then it works exactly the way you would expect: everyone works for 4 years and then leaves.
Imagine the East India Trading company advertising the TAM of the new emerging markets in the early 1600's. I bet the numbers compared to the old world would be enormous, and equally unbelievable.
IF, and yeah it's a gigantic really big IF... the future plays out the way people are envisioning it. It's future shock. Astroid mining, gigafabs, star ships launching hourly making space exploration cheap, satellite intelligence, physical AI. The world is going to be a completely different place IF SpaceX and the AI labs are successful. That TAM might be real. It's a literal moonshot, the stuff they're talking about sounds SciFi, but that's why the valuations sound SciFi.
That said, I would not invest anything into SpaceX that I wouldn't be willing to lose, and i personally would not invest until the lockups are free. Moonshots aren't in my risk profile.
This is such terrible history. Please Google things before making analogies.
The East India Companies (Dutch and English) didn't invent or discover the spice trade. They were created to leverage private capital to wrest control of the trade from the Iberian Union, an entity both Company's sovereigns were at war with.
The "TAM of the emerging markets" had already emerged both because of the Silk Route from Pax Mongolica and in the 1500s from the Portuguese Empire's Cartaz trade license monopolies enforced by their naval posts (feitorias) throughout Asia.
The TAM was incredibly apparent, Anglo-Dutch privateering during the war had seized multiple cargo laden ships along with trade route information.
I feel like you are intentially misrepresenting his comment to flex your history knowledge. English privateers captured the Madre de Deus. When they brought it to port, the cargo was worth roughly half the entire national treasury of England at the time. I think that is pretty equivelent to space mining, no?
We've been mining platnium and aluminum for all of humanity, so we know they exist, "The TAM was incredibly apparent". We know there are astroids worth 1 quadrillion dollars. So, if SpaceX has a 1% chance to capture that, thus the valuation.
a) If a single Portuguese treasure ship was laden with cargo of such magnitude, exactly how do you propose this is characterized as a TAM "discovered" by the East India companies?
b) Forget the asteroid, undersea mineral nodes are a total of 250 trillion tons worldwide, with an expected value of 233T USD. This is right here on Earth. Yet no country or company considers these minerals as a TAM, because the P/NAV makes this a net negative. This is the same reason any capex for Texas Shale oil will dry up if WTI falls below 60 for a period of time. No matter how many quintillions worth of oil are left in the shale reserves, if it costs more to extract than it will return, it isn't done. Arbitrary valuations mean nothing if they aren't economically viable.
c) Back to history. "Half the English crown treasury" sounds impressive, but the actual value was an estimated 500k pound sterling, upper bound 200M GBP in today's money. Quite a bit far away from 1 quadrillion. The English Treasury famously passed on a debt of 400k pound sterling to James 1 after the Spanish war, a contributing factor to a crisis between Crown and Parliament relations that eventually helped lead to the English Civil War. Suffice to say, a wartime treasury isn't very large.
Humour me for a bit. Let's assume we need to Hohmann transfer 1 quadrillion worth of minerals (30M metric tons) from a location in the asteroid belt to Earth. The DV necessary is 5000m/s, leading to a total energy calculation of 7.5 x 1e16 J. This is 78M tons of hydrolox fuel. 780k launches assuming 100 tons per launch (Starship). Are you honestly telling me that SpaceX is justified in adding asteroid mining returns to its TAM (1 quadrillion) because you believe it's economically viable to make 780,000 (lower bound) separate launches of fuel payloads from Earth to this asteroid? If ISRU is your claimed solution, then where is the Madre De Deus of this ISRU demonstration?
The most valuable substance in space rn is water. strip hydrogen and oxygen using solar power and place it in orbit. Instead of launching a Mars-bound rocket with 90% of its weight tied up in fuel, you launch it nearly empty, fly to an orbital fuel depot, fill up on asteroid-mined propellant, and go. Big business.
Or maybe you process insitu, nstead of a 30-million-ton rock, you return a compact, high-density capsule of pure platinum. A few tons of pure platinum group metals dropped into a stable Earth trajectory fits comfortably within existing down-mass logistics and could be worth billions.
Thirdly, (prolly the worst of all the options), you place metal slag in orbit and you drop it on peoples heads. Now you've got rods of god that any military will pay billions for.
Edit:
Lastly, once you control a vast amount of stratigically important material they would instantly become a sovereign-level geopolitical superpower. The moment a private entity controls a constellation capable of dropping kinetic rods anywhere on Earth within 15 minutes, every terrestrial military becomes fucked overnight. The line between industrial infrastructure and planetary hegemony completely disappears. ya gotta price that in somehow. Then... we really have realised the vision laid out in the book "the soverign individual"
Edit 2: >where is the Madre De Deus of this ISRU demonstration
If everyone could do it, they would have done it. The Madre De Deus was an absolute marvel of engineering at the time. the english knew the portuguese were selling all the stuff from the colonies, but they didn't really know how freakin' profitable it actually was. The Dues was freakin' massive. AKA Starship.
Let’s do a thought experiment. let’s say you find and can mine an asteroid full of rare earth minerals worth a quadrillion dollars, and you then flood the market with your yield. is this asteroid’s yield still worth a quadrillion dollars?
yeah, you hoard it like debiers and use it to build the city of elon, mega factories, or a giant platnium statue of trump to gain politcal favor. You def don't dump it on the market.
The TAM that spacex quotes has nothing to do with Space. It is all AI. As in grok. This is a joke, and people that believe it are just providing exit liquidity.
AI hype is full of bad total addressable market hypotheses. There just isn't $26T of money to buy this stuff.
Patrick Mackenzie, at the time at Stripe, about 10 years ago, talked about how Stripe's growth depended on growing the actual internet. Not just growing their market share within it.
Google, Meta, etc, have projects dedicated to getting more people online, because when you have 2bn users, you kinda have to make more users.
If that were to become reality, and I won’t guess at the odds of that, is because it grew the world economic pie. Thinking about it in terms of the size of the pie today is treating it like a zero sum game. It’s a different game.
Based on the disclosed financials by SpaceX, they have made $48B in revenue since 2023 with a cumulative loss of -$41.3B.
And yet, based on stock price and market cap, they are worth about as much as Microsoft. Microsoft generated $80B of revenue and $31B in profits...per quarter last year. SpaceX will never, ever, generate $124B+ in annual profits.
Until the alians realize this and throw the drugs in the harbor, leading the a US vs alians war to open them up for free markets access. Not sure this can actually realize their ambition.
What about the compute deals? $3.1B a month for half their compute? Elon built $6B of rev in 122 days. What can he do now that he's got a $85B war chest of cash?
xAI was retardedly late to the party. When Elon Musk founded xAI in July 2023, competitors had already spent years buying up massive silicon footprints and locking down cloud partnerships.
OpenAI had a 3 year lead, anthropic 2.5 years, and google was building TPU's since 2015... what are you talking about?
First, many more people then just them buy these things and have ramped up their spending over the last 2 years.
And others if they were earlier, the explosion and demand of investment increased exponentially over the last 2 years, as we can see in the profit margins of the memory companies for example.
My claim was not that they were the first ever, but that Musk mobilized a huge amount of resources fairly quickly in a time where prices were exploding and before many others.
Early enough that compared to now he got a ok price and he can rent it more expensive then he bought it.
I'm not saying its not a reasonable business, but we are not asking 'can it be a business', we are asking can it be one of the most insane companies in world history. The suppliers have realized whats going on and they are racking in insane profit.
That's why all the Musk fans have realized that they need to scream 'Terafab'. But that is a fanciful concept and unprecedented in history of semiconductors.
>we are asking can it be one of the most insane companies in world history.
"So we went from, he just got lucky cause he got in early", to, "okay, he wasn't early he was just really fast", to, "the suppliers now see he's making a ton of money so they are going to raise prices" to, "terafab is impossible", QED?
You seem to misunderstand how tier 1 capital allocation actually works. You don't get fucked by your suppliers when you are adding $B's to their bottom line. XaI never "Stopped" purchasing GPU's in 2023. They signed multi-year long deals for as much hardware as nvidia, supermicro, dell and micron/skhynix can provide. They are the last to get squeezed. They never stopped adding on to Collosus 2, and they are actively building collosus 3.
Lastly, nothing is "Fanciful" about terafab. ASML makes the machines, Intel provides the cleanroom tech, and 14a pipeline... these are known variables that need to be tackled. Tesla and SpaceX are world renowned, high throughput manufacturing facilities with the ability to make lines more profitable than their competitors. .. no one is claiming he's building a TSMC-style silicon foundry. They are building a massive compute fabrication/clustering footprint. It is a formalized $55B industrial plan filed in Texas, backed by a hard partnerships.
They have quite short cancellation notices and there is a question why do you need to rent it if you are an ai company? How is it grok not running the world using that compute.. And possible cancellation will erase a large income stream, think when gpus get old or sooner when Google does not need it anymore.
an RTX 3090 or older Quadro/Tesla enterprise cards are still selling/ renting... 5 year depreciation is only for tax purposes.
Anthropic is not building their own datacenters... They are mainly giving equity for colocation with AWS. I cant find one example of them ground breaking on anything... Can you give me a source?
yeah, and power is not availabe at capacity until late 2029? They haven't even signed anything, the 'deal' is in principle and BP has been actively trying to challenge it because they want the same power for a hydrogen facility.
So, to circle back to my original point, Anthropic has not broken ground on actively building any datacenters.
Enron was never audacious enough betting every US man, woman, and child will spend $28k/year on their generally nonprofitable business with one exception -- Starlink.
Patrick Boyle covered the SPCX trajectory fairly well... =3
That's the number AI boosters are spreading too "The TAM for AI is all of humanity - this includes every person and every company. So, imagine this huge pile of future revenue." I agree that TAM is likely huge but is SpaceX most suitable to capture that TAM? Unlikely. But for now everyone wants in on the AI hype train and FOMO of losing out one any company in the AI space.
In the long term most markets are duopoly with small competitors. And personally I see OpenAI and Anthropic duking it out rather than SpaceX.
I know this is just nonsense wish thinking, but apparently the investors disagree and I have zero clue when they will also stop giving Musk their money.
I think the bump since IPO can be explained at least partially by low float not meeting demand. I've seen a lot of accounts from retail investors who entered the lotteries saying they only got a small fraction of what they wanted, hence demand is kept artificially high. Probably intentionally, since it essentially allows the optimists to dictate the price.
(edit: This is not at all unique to spacex, of course, but given the nature of Musk's companies and their "fans" it's logical that they would employ this strategy. They are also doing a staggered unlock to avoid upsetting the market when insiders start dumping their shares.)
"investors" really over-sells who these people are. They're people with gambling apps on their phones who think SpaceX is great because of Musk's politics.
This is a stupid comparison, but Mojang/Minecraft was acquired for 2.5 billion in 2014.
Arguably the most popular video game of all time, which has brought joy to hundreds of millions of people for years and years, was valued at 1/20th of an AI startup that will soon disappear into irrelevance.
While Minecraft is just a game, I'd argue it has more societal value than Cursor. The way things are valued is nonsensical to me.
> While Minecraft is just a game, I'd argue it has more societal value than Cursor. The way things are valued is nonsensical to me.
Well it's because societal value is not profitability. Only question that matters is if Cursor can wind up worth more than 60b. Not even in raw revenue so much as ability to keep shilling the same story.
I'm not sure whether Elon's interests are in the model. Maybe the existing distribution channel which is cursor's users?
The whole ai business models seems tough to comprehend. Huge amounts being passed from one player to the other.
Minecraft erases productivity. AI coding vastly increases it.
The economic value of a tool is in the economic value it can produce. The economic value that Minecraft can produce is limited to how much money 8 year olds can wheedle out of mom and dad. The economic value that Cursor can produce is probably in the trillions, if they play their cards right.
You're also forgetting that money is worth a lot less today than it was in 2014. $60B today is $42B in 2014 money. Still 17x that of Mojang, but not the 24x that it appears at first glance.
It's more than just a game. It's a bespoke social network. It's a merchandise generator. It produced an entire hollywood movie. It's become a cultural reference.
It's been around for about 17 years, and if anything is more popular now than ever. So even if lots of kids grow out of it, there are new ones being born every minute.
I imagine at least part of the reason this doesn't make sense to you is because you're not a decision maker working at that scale. It made sense to someone at a high level. The calculations that person is doing is likely different than yours.
This is also why it's difficult to make money selling to consumers. They run different calculations than enterprise buyers.
There’s quite a few historical cases of mergers acquisitions being poorly valued aol time warner for example. So maybe v the calculus is different from the execs perspective but it’s not necessarily any better.
Idling? Where do you get your news? He's leased a ton of capacity to everybody else. Makes little difference to Musk whether Grok is using them or someone else is paying him to use them. We all know from AWS's profitability that renting servers to other people is great business. You get a check every month (often from other people's VC money) as opposed to having to figure out how to monetize.
I see it more like McDonald's can't produce burgers fast enough because of a meat shortage, but Burger King has their own supply chain that allows them to supply extra meat. McDonalds now needs to buy patties, possibly at a higher price than they usually pay, from BK just to be able to meet their burger demand.
Is BK in a bad position? They get to run their beef plant at 100% utilization regardless of how good they're doing at marketing hamburgers, and they can use that money to keep improving their hamburger quality or marketing.
My wonder is if anybody would buy anything if value == price.
The value of water to a person dying from dehydration is infinite compared to someone who's adequately hydrated. By that logic, this bottle of water is worth every possession, good thing, in the present and in perpetuity because that is the opportunity cost.
Nobody would choose $50 million in cash to reject the water because the money only has value if you take the water.
But hypothetically, if the value and the price could be finitely defined, a "fair trade": let's say 10 apples and 1 watermelon are each worth 10 utility units. Price still can't equal the value. We don't eat utility units. The watermelon inherently provides a different value than the apples. An apple isn't a substitute for a watermelon.
I think this is my long winded way to conclude that trying to compare everything is apples and oranges, but somehow we still try to give it a dollar amount.
> The value of water to a person dying from dehydration is infinite compared to someone who's adequately hydrated.
But most people are adequately hydrated in steady state, and to them the value of water is not that high. Since today it is necessary to sell an item for the same price to everyone, selling water for infinite money would lose nearly all customers.
(Not to mention that it’s possible to acquire water from natural sources, for free)
Somewhere between ~15-75+ billion USD. Higher end if you want to include the hardware R&D/etc, which is arguably core to their metaverse concept. I love the idea that they don't even have a product _called_ metaverse. Most of the VR content developer acquisitions costs by Occulus Labs are undisclosed.
> "In its IPO filing, the company had said Cursor's access to developers' data, including coding requests and design decisions, could help improve its AI models such as Grok."
They're all stealing your IP and selling it back to your competitors in the form of tokens.
> "I want LLMs to code for me, but I want them to be trained on other people's code, not mine, duh".
Who ever said that? Have you actually heard that from your fellow programmers in real life?
If the code I wrote actually made even the slightest discernible difference in LLMs I'd be so honored. But it won't happen, as it's just 0.00001% of all the training data.
Tragedy of the Commons is not an analogy, it is an inevitable result of a large fraction of participants in a coordination game defecting due to perceived individual advantage.
Sounds good? They can pay for code they want to train on. There are plenty of companies sending me offers to code training materials for them for $50-100/hr. Don’t expect to charge me an arm and a leg for inference and then also train on my code.
There are already opt out buttons for training in Cursor and Claude Code… if you don’t want it then turn it off. If it was worth enough money to them they would offer a monetary incentive like discounts but none of them have yet
Interesting how our generation which grew up using Napster now has so many intellectual property extremists. By this logic, even humming a tune you heard on the radio is theft.
> I have absolutely zero interest in free. I honestly don't think I'm even remotely in the same demographic as people using free tiers / models. I want to pay. I don't want my data used for training...
They want to use LLMs trained on others code but don't want to contribute with their own.
It makes sense from a business perspective-SaaS firms value the ability of coding agents to accelerate development, but also worry the models will learn the secret sauce of their business and destroy its moat. So their desire to contractually exclude training on their data has some logic to it.
(Disclaimer: Not speaking for or about my current employer, just a general industry observation.)
I don't really use LLMs myself, but if someone wants to have any kind of software business then having the models trained on their products isn't ideal.
Cursor users are willfully providing it by using their product. Not unlike uploading a personal photo or video to social media -- that's not yours anymore. You gave all rights away when you put it on their servers.
> Not unlike uploading a personal photo or video to social media -- that's not yours anymore. You gave all rights away when you put it on their servers.
99% of users are unaware of that, so it’s false to say they "are willfully providing it".
Do you think that everyone using it and their employers are aware that they are giving their competition the ability to copy straight from their codebase when they ask it to replicate their product?
presumably they are subject to audits. i work at a tech company that has to deal with data compliance and we take it incredibly seriously and have automated systems tracking, tagging, and deleting data everywhere.
> Not unlike uploading a personal photo or video to social media
I've granted them a limited license to use it.
> that's not yours anymore.
Not by any definition in the contract or in law is this true.
> You gave all rights away when you put it on their servers.
I gave away some rights. I also got something in return. Attention. And at the end of the day I'm completely entitled to turn around and sell copies of this work for profit. The only thing I can't do is sell an /exclusive/ license because that is no longer available.
None of this provides any implication for people who upload code to their own websites. Which these rapacious LLMs bots happily index, sometimes to the extent they actually crush the site, or create unusual costs for the owner.
Finally none of these LLM companies tell you where the source came from. Whether it is copyrighted, whether ownership rights are retained, or whether the code can be used publicly or not, and if so, which license it's covered by.
You're using the lens of social media contracts to understand something far larger and more important. It's lead you to some bizarre conclusions and huge oversites.
What if someone steals my work and then uploads it to facebook and claims it as their own? Do the rights no longer exist because it got uploaded to Meta?
I work at a 20k-employee company with footprint in tech. A couple of weeks ago a decision was made to not renew Cursor license. This week it was made public and has been greeted with groans. Some employees have been publicly saying how ineffective and unproductive they are going to become due to this decision- which actually strikes me as really naive.
Personally I like to use a stable IDE. I used Cursor for a couple of days and then went back to VS Code, largely due to Cursor pushing agentic first approach with V3 update.
Indeed, according to Cursor leaderboard there were folks committing 200k-300k lines of code every month. And here I am intentionally slowing myself down manually approving every AI generated code change and terminal command.
> They're all stealing your IP and selling it back to your competitors in the form of tokens.
The users of those tools are stealing too. The model is trained on free software licensed under specific terms and the output of a prompt will strip those licenses and their terms.
It's not theft in the same way a con artist convinces you to give him all your money. The people using it or their employers just don't realize any competitor will be able to ask the LLM to replicate their product and it will copy the codebase they uploaded to them.
You could just test it for yourself if you're actually interested in finding out (assuming you are a decent programmer, but your bio suggests you have no technical expertise in the area).
First I never made any claims about myself or appealed to my authority, that was you. I still don't see any evidence of any technical expertise in ML or in programming anything more complicated than python scripts. Maybe you weren't aware but often on the internet people make false claims about their expertise on something. There's thousands of "Visionary leader driving ethical AI innovation and strategic growth" CEOs that know next to nothing about the actual technology.
Anyway like I said you can just test it out yourself and find out that I'm correct. Every skilled programmer already knows this and can predict what kind of complexity an LLM won't be able to handle. And anyone working on LLMs should know that they are completely dependent on their training data. The entire scaling hypothesis was based on this.
There's open source ML&DL projects dating back 12 years on my github, I don't know what you're on about.
Regardless, your claim is "LLMs being able to one-shot any higher-order complexity is entirely dependent on it already being in the training data." which is currently unfalsifiable, and known by every AI researcher to be most likely wrong. You've also said it's "been demonstrated". So stop wasting people's time and link to the demonstration instead of hand-waving a "do it yourself you're smart".
Good lord, it's like talking to a bloody climate change denier, I swear.
I have neural net applications I programmed entirely myself from 16 years ago and all I see from you are some python scripts so I guess I win this stupid game I wasn't even trying to play.
You haven't made a single technical point outside your bloviated claims to authority. I can safely assume you're a fraud because this is exactly how frauds speak. Actual scientists and engineers don't argue from authority they go and test the hypothesis for themselves, the fact that you balk at my suggestion to do this is amusing.
I said you can do a basic test because this is the best way to see it directly for yourself. It's very easy to do especially for an eminent machine learning visionary leader as yourself. You ask the LLM to produce two apps of similar complexity and technical challenge from a software perspective, one that is already in its training data and one that isn't and see which its more successful at. This isn't some controversial take, nor is it "unfalsifiable".
There's also hundreds of benchmarks demonstrating where the limitations are for LLMs, or you can study the progression of LLMs in mathematics and where the gains have been made and see that this also agrees with me. You can watch Chris Hay's videos demonstrating exactly how LLMs perform math layer by layer. Why is everyone using LLMs for search? Because it's an extremely efficient compression of all its training data. Did they figure out the Studio Ghibli art-style all on their own spontaneausly? No, they were trained on Studio Ghibli content. There's so many ways to come to this conclusion. But you seem to be too busy sniffing your own farts to be interested in learning anything about the field though.
This pointless test you're suggesting is a few countries removed from your initial suggestion, which was as a reminder: "The people using it or their employers just don't realize any competitor will be able to ask the LLM to replicate their product and it will copy the codebase they uploaded to them" (which, yes, is unfalsifiable given the search space)
I mean, are you seriously trying to back off that original ridiculous claim into a "code in the training data is more likely to appear in the output than code that isn't"? And I'm the fraud? As I said before, get over yourself.
Yes a test showing an LLM can reproduce an app in its training data and not an equivalent complexity app that is not in its training data is equivalent to proving the statement I made that having your codebase trained on will allow a competitor to copy your product.
My whole team was on cursor for a few months. I enjoyed using it and thought it was the most complete of the agentic coding tools I tried. The thing that got me was the cost. I was switching between Opus and GPT 5.x and was spending anywhere between $500-1000/month. I was using a relatively normal workflow, paste in ticket, plan, execute with dumb sub agents, have the ai test and competing model to validate. The business got uncomfortable with the cost when everyone started doing the same so they switched us to Claude code since it has better cost controls. So far it looks like we won't even touch the $100/month plan and some people would be ok on the $20 plan. Anthropics usage limits is a consistent source of complaint on here but I've found them to be moderately generous in comparison to cursor. Cursor also charges a $.25Mtok premium for 'routing' no matter what model you choose. 5% increase for frontier models but when you're using haiku on sub agents that's a 50% cost increase. Composer is solid but if you don't have deep pockets it's the only feasible model on their platform because of how they bill it. Being an all in one editor/agent is nice but if you're in a language like c# or Java you're already swapping back and forth with a real IDE anyway.
Our experience has been the direct opposite. The way to get the most out of cursor was to leave it on auto - we saw it average around 3.9 cents per request under the old contract (per-seat pricing) and more like 39 cents per request under the new (single pooled cost for everyone). Composer came in a little higher, more like 50c/r, while the claude models were up into the dollars. Meanwhile, if you use Claude-the-app, there is _no_ cheap model and the default switched to Opus in April, resulting in increased costs across the board.
We have both Claude and Cursor here, as well as agents running GPT, things in AWS Bedrock, etc and its my team handle the bills...when people use Cursor on auto, costs are under control, but there's always a dozen or so whale users who'll switch models manually and blow through the budget like it's not there.
Another thing: "better cost controls". There was for example no way for us to disable Fable in Claude, but we could in Cursor. Again, the opposite experience.
For the kind of ML work I do Cursor's "Auto" was too unreliable. It would frequently try to solve relatively difficult tasks using Composer, when it should have been routing to a true frontier model. Annoyingly, doesn't tell you what model it's decided your task deserves, so I often wasted time working through a problem with it, only to realise I was on a dumb model. Then I revert all the work, switch to an expensive frontier model and pay expensive API prices to actually solve the task.
All that is to say, model selection is the main control we have over quality. Giving it up in the name of cost saving will bite you in the long run, especially when Claude Code still has such good plan pricing.
I also had a problem with Cursor auto switching on fast mode which I would never use otherwise.
Cursor really needs a stronger model than Composer 2 which is the root of the problem. Hopefully they will put out something better now that they have better datacenter access with SpaceX.
Well, auto can produce garbage code. I'd rather be able to select the model manually myself and so far it's much more economical to give people Claude Code and Codex subscriptions as those are subsidized per seat plans.
More precisely -> the competitors heavily subsidize which causes cursor to feel expensive.
> Recently, we purchased one of each Anthropic/OpenAI subscription plan and randomly ran long horizon coding tasks until we exhausted the weekly limit. It's widely believed that a $200/month plan maxes out at ~$2000/month worth of tokens (assuming API pricing). However, we found that the subscriptions are actually far more generous. (2/4)
Just setup my team with Cursor. They actually got back to me on an enterprise plan (Claude keeps ignoring me). Cloud Agents have been great for keeping multiple streams going at the same time. Adding in computer use has been great for actually testing out features and showing they work for PRs. Bugbot so far has been the best AI reviewer I've tested. Composer 2.5 is great, though still using Opus for planning.
I can do most of this with Claude Code, but there's definitely a cost in maintaining it for the whole team.
It's my primary. Claude Code for personal stuff on the weekends. I really just prefer the GUI of having the changes easily highlighted. If I can get something to apply that with Claude Code or Codex or OpenCode or whatever, I'd swap over without thinking.
Same here - their UI/UX seems to serve my workflows/habits the best. And it's strange that no one else seems to be delivering a compatible experience for me. I'd prefer to move away from Cursor after this acquisition.
The Claude Code extension in VSCode is similar enough. I used to use Cursor, but now I use VSCode+CC to take advantage of the better pricing CC offers for always-on frontier. I found Cursor's affordable "Auto" option unreliable, often wasting my time with dumb models.
As in the changes its made? You can do that on Claude Code but by having it do it on a worktree and just checking out that worktree. Then you can see the change file(s) in your preferred IDE
I stopped using Cursor because of how terribly optimised it is (worse than VSCode despite being a fork). It would routinely take up 50% of the CPU resources on my MacBook M4 and gigabytes of RAM for absolutely no reason.
I switched to Zed, and I'm never going back to Electron/non-native IDEs.
Yes, I use it as my daily driver at Discourse, the Cursor Tab autocompletion is still the best AI-based editor autocomplete I have used. I switched to nvim earlier in the year because there was no real great AI integrations that didn't feel cumbersome, Cursor in vim mode is a much nicer experience, I like adding lines of code to agent context and so on. All of the extensions from VSCode, LSP integration, and so on are really nice.
That being said I do a lot of work in Codex or Claude now (they all feel pretty much the same to me), and use Claude for manual code writing and tweaks that I feel would be unnecessary for agents to do, or just when I am writing code that I enjoy writing, typically when exploring a new problem I'm interested in, not everything needs to be done at 1000% speed by a robot.
Though I will likely reconsider this now that Elon will own it, either moving to VSCode or to Zed long-term.
Still do. Composer 2.5 is a beast. But even with Opus (and Fable for a few days) their harness is many times faster. The main reason for me to use CC is the $200 subsidized pro max plan.
Also their computer use in the cloud agents (when it works) is a game changer. No need to keep your laptop open / get a Mac mini if it runs in the cloud.
I was on Claude Code the past year, now I use chinese models, but I've used Cursor and they have an ok pricing offering today because of their mix of sota models with usage based pricing along with their Kimi based Composer model with generous limits. I think it makes a lot of sense for the enterprise market, which is the real moat, and not the capabilities/features of the forked ide or app/tui/github bot anyone can come up with today.
I much prefer Cursor to Claude Code. Claude Code hides too much of what it's doing from me (can't monitor the thinking output, can't see output from in-progress commands being run.
I don't even use the IDE -- just the Agent Window interface. I also really like Composer 2.5.
That's good for you but the demand is so low SpaceX is leasing a considerable amount of its compute to Google and Anthropic to compensate. Regardless of whether it is useful, it is clearly not popular
I use Cursor for coding. I like to review the changes via the UI. Plan mode is also really strong in Cursor. It bugs me less about needing to search through files and basic coding tasks. I find it also saves the company a ton of money compared to Claude, Claude burns through tokens with no regard.
I typically use Claude for interacting with MCPs and skills to operate on live systems.
Plenty of enterprises are still using Cursor, though they are facing plenty of pressure because Anthropic and OpenAI bundle Claude Code and Codex which can make it hard to justify an additional license for a third-party harness (why spend that money there when you can buy the underlying tokens instead).
It's still my primary at work (data engineering/platform engineering), running a mix of GPT-5.5 and Composer 2.5. We also have Claude Code subscriptions. I find myself preferring Cursor for most tasks.
70% of the time I use AI agents on a pretty tight leash. I often reject edits and ask it to change things. The IDE integration is really efficient for this workflow compared to Claude (yes, I've tried the Claude extensions).
Autocomplete is still the best available (I've tried both Copilot and Zed); though admittedly it's not as important as it was circa last year.
For the 30% simpler or very well-specced tasks their cloud agents are last I compared way better than Claude Code's/Codex's version of the concept. @cursor for quick fixes in Slack works quite well. Don't get me wrong, it's still quite under-documented, but the others are worse. The integrations with linear, Slack and github are well executed
Composer 2.5 is really fast in their harness at code search/explanation/Q&A tasks (much faster than Sonnet/Opus). It's also really good at debugging, very proactive compared to other models in the same size/prices class IMO. Just due to the speed I actually prefer it to almost any other model for these tasks. I suspect at least some of this may be due to the harness and good codebase indexing.
I don't know why people are down on the Cursor harness. It's good. The main advantage of Claude Code/Codex are the token subsidies; but according to their dashboard I am costing my employer between $100-200/month on Cursor, so the overall price is comparable and only narrowing now that Anthropic is switching many enterprises to API usage.
I also don't understand the people complaining about VSCode bloat and in the same sentence praising Claude Code. Claude Code often uses MORE RAM than Cursor, has a super unstable UI (on my home machine there is input lag when typing ANYTHING in Claude Code) and the desktop app version of CC barely works. The Codex TUI is genuinely nice and snappy, on the other hand.
I'm still on their old 500 requests/month plan and the value is simply unbeatable for $20. I've been able to use agents without worrying about usage for my job and personal projects paired with the $20 codex sub and I dread the day when they finally get rid of the requests plan.
I use Cursor with the Claude code plugin because Cursors autocomplete is really good and I like the way Cursor is set up. But it’s definitely a UX downgrade using a plugin instead of the builtin Cursor agent stuff which I do miss. Claude with Max are just a safer bet economically these days than a cursor plan.
I except that model and pricing gap to narrow over time though
I can only stand to use the cursor CLI at this point, but it's usable I like to be able to switch models when one gets stupid or guard-railed, on the pro plan you get so much usage in the composer 2.5 pool you could just use that your whole sub if you wanted.
Development is moving away from the IDE to agentic long running workers. I've been using their SDK in this mode - which then forces you to use cursor as model provider. I use a mix of harnesses for different types of agentic tasks and Cursor gets the best results.
I don’t see how Cursor differs in that regard. They have an agent management system since Cursor 3 and some feature that offloads agents to the background or cloud which I haven’t used yet. I always just keep chat windows open if it’s running because I’m always watching that stuff closely.
Seems more like an advanced/niche feature for people who go really hard into LLMs IMO
Earlier this year I had used it because I would rather have a IDE-like exp and be able to actually look at the code. However, recently switched to using claude code VS code extension and it's basically the same thing (plus at Amazon we can only use Claude Code)
I use Cursor, but funny enough it's 98% just using the codex plugin - I kept cursor around on the grand fathered $20 / 500 requests plan, if they un-grandfather me or things change too much I'll zip over to vscode.
I use Cursor and it’s been fine. I write a lot of code manually too, so I liked the tight integration with VSCode, my daily driver for about a decade. I used to use Vim, so I’ve “discovered the terminal” a long time ago.
The people steering the agents are the ones producing low quality code. I see little correlation outside of that.
I don’t understand this take. In my company a lot of the Claude Coders seem to be very uninterested or unaware of the code they are producing, while I in Cursor usually click ”Keep”/”Undo” on specific code blocks (with little edits) or sometimes the whole file at once if it’s a low risk part of the codebase. I fail to see how this workflow produces inferior code vs shooting in the blind and maybe skimming a huge diff in one go.
I’m running Opus through Cursor like I would in Claude code, so make sure you’re not comparing different models.
But the core difference for me is I can easily see the code changing in cursor, but not in Claude code. Bigger tasks, I’ll have to drop into other tools to see what Claude is doing along the way. I don’t like that. I like everything being in the same spot
i use literally all of them, they're all pretty good. codex is the lightest on my machine which is nice but less featureful - the others are all pretty at parity (minus a few CC-only bells & whistles)
$60B is a huge price, but buying Cursor gives Musk something xAI has struggled to build: a popular coding product with real developer and enterprise adoption. It may be the fastest way to catch up in AI coding. The real question is whether SpaceX ownership improves Cursor or drives its users away.
> The real question is whether SpaceX ownership improves Cursor or drives its users away.
I know I'll sound hyperbolic but I'm deeply skeptical of the way anything Musk-owned is going to treat private data. I think he wouldn't hesitate to dig into it if it were to his benefit, even if there was an agreement against it. For that reason alone it makes Cursor look worse to me.
A) With great power comes great responsibility + making the world better through eradication of diseases, heightened food and economic security, improvements to the stability of nations, access to shared research and information, and increased personal agency makes it all better for Americans too.
B) Given you probably find (A) uncompelling, you should familiarize yourself with the concept of "soft power" and try to learn a bit about how relatively inexpensive it's been for the US to have an outsized influence around the world over the past many decades (outside of stupid, unnecessary wars, of course) and the benefits that's yielded for our economic and geopolitical interests across the board. If you'd prefer to not though, don't worry -- provided you're around for a couple more decades, you're going to get a crash course in the alternative.
> The real question is whether SpaceX ownership improves Cursor or drives its users away.
I can't speak for anyone else but I wont be renewing my sub. Funding anything Musk related isn't exactly high up on my list of desires, and theres ample alternatives out there.
My sense is that enterprises are extremely cautious. They like everything that is already common and hr friendly. They abhor anything that might be seen as divisive and controversial. That's why they're currently going with Anthropic and not Openai or Xai or anything Chinese. It's the smaller actors that are using everything but Anthropic. Anthropic got that safe enterprise bland vibe. The only pr trouble Anthropic is in is with saying no to the military, which just makes them even more enterprise safe. Meanwhile Sam Altman and Elon are out there freaking out the enterprises almost every day it seems like.
I think it'd be an enormous endorsement of Cursor/xAI and proof of improvement if SpaceX started using it to code the mission critical software running on Falcon 9. Which other AI company can say their models powered a rocket launch?
(mostly /s but I know I'd give it another look if it was that good)
> Which other AI company can say their models powered a rocket launch?
Honestly, probably all of them. I imagine those coders are using all the tools they have available and are using Claude and ChatGPT as well as internal tools.
I can't imagine that any of the cautious companies or ones with their ear to the ground are going to want to cozy up to cursor with this acquisition; I'd suspect we'd see some exodus as well given the relationship to Musk.
I've tried most of the tools out there but have used cursor most consistently. Sure, some of the UI quirks get in the way sometimes, but I've found its auto complete predictions to be unparalleled. More importantly, these days I mainly use its Ask mode, Plan mode, and Agent mode. I like that I can use Opus via subscription pricing without Claude Code's wild and buggy harness. And I find cursor's plan mode to perform better than Claude's, but that may just be my personal preferences. I know cursor stopped being the cool thing a few months back, but I genuinely feel most effective with it!
But I'll stop using it now, for the same reason I wouldn't buy a Tesla, or support that maniac in any other way. And I'm sad about that :(
I used Cursor for a couple of months like a year ago, and I found it to be good, but I preferred the CLI experience and switched to Claude Code. "Claude Code's wild and buggy harness": not sure when you tried it, but I've been using it daily for the last couple of months, and okay there were some quirks here and there, but its stable and very usable.
To be clear, you can totally use a claude code subscription in Zed. You just risk a chance that they're ban you and your subscription for it. Or you have to pay API pricing. Who's got money for that though, shit.
Signed up for Zed upon seeing this, signed in with GitHub, it failed instantly. Emailed support, they asked me to prove my ID because I was flagged as a bot. I immediately revoked access.
Fairly certain the support reply was from a bot, too. Meh.
Cursor has $4 billion annual revenue rate so $60b is 15 years of future cashflows.
That's not crazy because if past predicts the future, that revenue will grow quickly. At $8 billion/year it's just 7.5 years, which is a reasonable investment.
This is a linear regression relying on a couple years of data to predict 15 years in the future and I don't believe that the valuation is made on this basis.
It may be that spaceX is buying an operation that would realistically take 5 months and 100 million to copy in-house for 60B because the worry is that waiting 5 months might cost that much in some sort of lost opportunity. It also might be that in any negotiation SpaceX is viewed as incredibly cash-rich and so anything can be sold to them for inflated prices.
I really don't understand these companies valuations it seems like boardrooms everywhere are in a constant state of panic that they'll lose it all if they aren't growing a breakneck pace constantly.
Many made the same prediction about Twitter, and it seems to be more or less the same or higher activity than before, and Bluesky is continuing to rapidly decline (-5% DAUs per quarter).
I wouldn't bet on Musk, but I DEFINITELY wouldn't bet against him. Anyone betting against him over the last 10 years has been viciously smoked (many short-selling hedge funds got wiped out completely).
> I wouldn't bet on Musk, but I DEFINITELY wouldn't bet against him. Anyone betting against him over the last 10 years has been viciously smoked (many short-selling hedge funds got wiped out completely).
In terms of the stock market, definitely. Honestly though, all those people who said self-driving wouldn't be solved by now, that Tesla didn't have a great moat and that the Boring Company was profoundly stupid were in fact correct.
> Many made the same prediction about Twitter, and it seems to be more or less the same or higher activity than before
Where are you getting that? There's not a single piece of data out there that will tell you Twitter has increased in users. Not only have they visibly dropped in users, but it's becoming increasingly clear the site is Astroturfed with bots.
The choice is not Twitter or Bluesky. Most people moved on to TikTok. I don't know a single person who uses Twitter.
Why would that be germane to the discussion? Are you under the impression people are hardwired to use specific kinds of social media?
The thread is about Twitter supposedly *growing* in size recently, and its only evidence was Bluesky's userbase. That's just silly. I know half a dozen people who left Twitter for TikTok. They don't care about the slight differences in these platforms, they are all algorithmically driven platforms where people comment under text and video. They care more about what audience is cultivated and what is forced on them.
> Many made the same prediction about Twitter, and it seems to be more or less the same or higher activity than before, and Bluesky is continuing to rapidly decline
Twitter did lose significant revenues. It came ahead on profit because it aggressively cut costs which is an interesting thing to do for Musk given SpaceX. Maybe he learned that profitability is over-rated?
I looked the other day, the feed got the same 4 postings reposted 3-5 times. A dozen posts talking about people getting banned, doxed and fired for talking about you know what.
I found 1-2 sort of interesting posts but not interesting enough to make it worth it.
It was fun for a while blogging with grok back when it was free.
Kind of interesting that engagement is zero on all of my postings. Not that I care for it but it shows quality isn't measured by users. The system prefer people read about users getting banned and that same video again and again. lol
With rss I just sort by date and eliminate duplicates. Not very hard and dramatically more interesting.
That revenue number is almost meaningless, since they give out tokens at a loss. Especially with Composer 2.5 tokens are sold at a steep loss. They could certainly grow to $8 billion/year, with this negative revenue / heavily subsidized subs, but what will happen if Cursor decide to be profitable, or maybe to even just break even?
It's a crazy number especially since Cursor feels kinda dead. Few thoughts from the other side:
- xAI needs the coding related data to compete with Claude Code and Codex
- Recent progress with Composer 2.5 seems promising given the size
- The may get a comeback on the smaller than Enterprise battle field now that the other two got so expensive
- The way that Elon set up this entire process was quite genius. They locked in this option before, and now after the gains through the IPO, it feels almost like a discount, lol
I wonder what happens to fireworks ai, who provided the infra to train and serve composer 2, cursor was their largest customer, and they're probably loosing it.
I could see it being a talent / first mover acquisition. I’m bullish on harnesses, but I think there’s still a very long road to get to something that is stable and relatively optimal - harness user experience is pretty trash tier right now imo.
My hot take is that it will probably be like the OS landscape:
- Some established enterprise option (Windows)
- Quality secondary option for professionals (OS X)
- Super users / nerds / tinkerers (Unix flavors)
Here's a shower thought. BTC essential is worth $70k solely through the power of memes. Can TSLA and SPCX remain overvalued (relative to the revenue of their respective underlying assets) forever through the power of memes?
Intuitively, it seems to me that there must necessarily be some kind of upper limit, but I'm not convincing myself. These speculative assets are only attractive as long as the price keeps inflating. But that can only happen if there is more and more demand. So it's basically a bet that there is an average amount of retail investors (I assume it's mostly retail investors but I could be wrong) that consistently put a percentage of their income into these speculative assets. Can this be maintained forever?
The saying "The market can remain irrational longer than you can remain solvent" exists for a reason.
In short, the answer to your literal question is "no" because nothing remains forever in this world. The practical answer is "yes" because the TSLA stock has been irrational for years already and it shows no sign of stopping.
Or the way I would think about it is (I think this was a Krushchev quote): You don't convince people there is no imaginary river, you convince people that you built an imaginary bridge across the imaginary river.
So you are never going to convince people that Tesla won't make a fortune on humanoid robots or SpaceX won't colonize Mars. The fever will break when people decide flying robots (or whatever) are a bigger business then humanoid robots or some other company than Space X will colonize the ocean before Mars. You aren't going to convince people to price these things reasonably but eventually the heat will wear off.
That saying assumes a rational market, and while there is some evidence that the average human behavior trends towards the rational there are plenty of evidence against individual behavior being rational (see homo economicus).
As more and more wealth get distrubuted to fewer and fewer hands, and as fewer and fewer extremely rich individuals control more and more of the market, My gut feeling is that if the market was ever rational (which btw. I am not entirely convinced of) that very much no longer holds true.
The before ones have had outside influence which forces the irrational episodes to pass. A spectacular collapse is usually fallowed by mass protests and labor actions, strikes etc. which forces the market to change its behavior. Most of the time these will result in a change in government policy which imposes regulation onto the market, and only then does the market go back to being rational.
The reason is that people keep making the same mistake, because people are not very good at assessing high-risk, high-reward projects.
Only a real deadline/delivery failure can wake people up, and only if they haven't pivoted their dream to something else, and only if you don't have other people knowing it's a scam, but willing to prop up the stock price because they are highly invested.
Scams, like cancer, are real; they survive, grow and defend themselves using the same mechanisms (laws, advisors, promotors) as ordinary investments/tissue, until they kill the patient -- so the best scams target the largest unkillable patients and enlist the broadest and deepest range of self-interested insiders as their defense.
It's beautiful, if you really think about it, as a tragic example of the worst of capitalism.
The thing is fundamentals really don't matter. TSLA and SPCX aren't paying dividends so there's no real performance they have to hit, no one is going to miss a dividend payment and dump the stock. The Elong vibes can carry it as long as people keep smoking what he's selling
The real question is, when does that run out of steam? When do we wake up to the charade that has built up around us? That's a much bigger thing than just Elon and his businesses. Like someone else said, when the next crisis/downturn/depression hits the house of cards will fall. Unfortunately it will hit all of us not just people in the meme stocks
I'm extremely cynical about the way the US government would react to any sort of financial crisis. I do not believe that they would not completely cave and bail out the AI companies and the monopolists if there is a downturn. And it's not that I don't trust the Republicans specifically. The whole political sphere is completely convinced that AI is a national security prerogative, and the cynical political atmosphere will equate national security with investor protection.
Let me append the saying a bit: The US government can remain irrational longer than you can stay solvent.
> If you exclude dividend paying stocks, then the entire stock market starts to look like a giant pyramid scheme casino.
Stocks can start paying dividends in the future: MSFT did not in the past, and does now. AAPL did, stopped in the 1990s, and started doing so again.
You should be indifferent to the company's dividend scheme since it's the underlying business activity that drives total returns, and not its distribution policy. There is all sorts of magical thinking when it comes to dividends:
A pyramid scheme can run out of people to keep it going: the stock market is in a sense a 'savings scheme' for future consumption. Younger people work and turn their cash into savings, older people take their savings and turn it back into cash: as long as young people need to think about the future, and older people / retirees need to pay bills, there's a mechanism to maintain this cycle.
> A pyramid scheme can run out of people to keep it going
And then you describe how the secondary stock market requires 'fresh blood' to whom to sell stock to cash-out.
It's precisely a legalised pyramid scheme that always needs someone to come in at the bottom hold the bag to let someone else cash-out. In turn they need someone to come in 30 years later. That's exactly how a pyramid scheme works.
> In turn they need someone to come in 30 years later. That's exactly how a pyramid scheme works.
The entire economy is a pyramid scheme: the expenses of some people (shelter, food, clothing, entertainment, etc) are the income of other people (landlords/mortgages/property taxes, farmers/grocers, etc). It's why, during economic downturns, personal virtues (saving) can become vices from macroeconomic perspective: if everyone is saving, no one is spending, and so producers/suppliers lose their income (and generally start laying people off, which causes more saving / less spending).
At any given point in time, if no one spends, then no one has income.
This was the 'innovation' of Keynes in the 1930s: use government spending to 'induce' demand to get the cycle going again:
For a stocks point of view: if no one is currently saving, then those that need income will lose it. At any given point there are folks who need to save/buy and those that need to spend/sell.
> A Keynesian beauty contest is a metaphorical beauty contest in which judges are rewarded for selecting the most popular choices among all judges, rather than those they may personally find the most attractive. This idea is often applied in financial markets, whereby investors could profit more by buying whichever stocks they think other investors will buy, rather than the stocks that have fundamentally the best value, because when other people buy a stock, they bid up the price, allowing an earlier investor to cash out with a profit, regardless of whether the price increases are supported by its fundamentals and theoretical arguments.
Plenty of folks may think these companies are garbage but are 'playing along' because it's not necessarily what they themselves think that is important, but what others think. You can make money in a bubble, even when it eventually pops. What we're seeing now is hardly new, either:
I generally check my portfolio once a year, in January, when I top things up when new contribution room becomes available with the new year. It's 'fun' to follow along with the gyrations and drama as things happen, but I don't sleep over it. If you're reasonably diversified you can generally weather storms and come out okay on the other side:
I don't have better ideas than you to do anything except "buy indexes".
Everyone is happy enough to give Elon (and others) more and more leverage to buy politically strategic companies (this is not that, this is probably just an ego buy for him, something to kill time because he can).
I was worried about him selling out (from an overall market and even index perspective when they were going to bend rules), but it looks like largely the whole situation is predicated on the idea that he can't or won't sell. I don't know how exposed the market is but it doesn't feel good.
No shit. That's why, even if it's an exaggeration to call the entire stock market a pyramid scheme, you can't justify the claim that it's entirely "underlying business activity that drives total returns". That's the real question (from which dividends are, yes, a distraction).
A lot of folks think the top ten stocks in the S&P 500 making up ~40% of the capitalization is bonkers, but they also make up ~40% of the net income share:
So from an earnings/income perspective, there appears to be a link between the two.
Perhaps worth noting that the US markets seem to (only?) outperform when tech is outperforming, with other US non-tech sectors basically performing the same as out countries' non-tech sectors:
No, there are a huge number of companies that are valued reasonably for their revenue / profitability / growth.
There's basically two stock markets: things valued on fundamentals and things valued on vibes.
And I don't think there's ever going to be a unified theory of value that can span both, because the former is quantitative and the latter is psychology.
True, and it's a common question that comes up: What's the point of owning a stock if it doesn't pay a dividend? How do you participate in the company's performance?
There are other ways for performance to translate to the investors directly. For example, if the company sells itself then all the shareholders will get that payout. Stock buybacks are a thing. And, as other commenters here have said, eventually the company may start paying a dividend
But, you're not entirely off-base in that you're just buying into the vibes of a company. It's just that most of the time (much of the time?) those vibes have been rooted in some semblance of rationality, that there was something of value behind the shares you're buying. That is definitely not universally true anymore
There's taxation reasons. Stock buybacks mean that investors get the usually lower capital gains tax rates instead of dividend income tax rates. On top of this, it also juices earnings per share.
Gamestop is not an Elon venture. There's plenty of irrational stock valuations out there, most of them just come back to Earth on a reasonable timeframe
> And being higher over 10 years has little to do with it if acts counter cyclical to stocks and other assets.
BTC has been called many things at many different times. It was originally a payment system:
> A purely peer-to-peer version of electronic cash would allow online
payments to be sent directly from one party to another without going through a
financial institution.
And it can still be used for that, however the transaction throughput is tiny, and so it became a store of value in essence: but it's kind of hard to be that when the value swings up and won so much. While "fiat" currency inflation is annoying, it is, generally, fairly predictable in most cases (<4%) and so you can plan ahead with regards to future value and purchase. The same is hardly true of BTC.
As the sibling says, the original value proposition of Bitcoin was that it would supplant Paypal and wire transfers. The store of value narrative was a post hoc rationalization by stakeholders when it became clear that it was technically unsuited to serve Internet-wide throughputs. Not only that. I sat in during the meetings around the time of the segwit transition; I think they were debating making the blocks bigger or something. It was amazing to see first-hand how such a relatively minor and technically necessary change could generate so much friction. That was when I realized Bitcoin was never going to evolve in a useful direction, and sevenish years on I see that I was correct, as the block structure is pretty much the same as the last time I saw it.
I mean the original thing I replied to. The idea that it was a stable or anti-cyclical store of value emerged after it became clear the payments stuff was not working.
There is some misunderstanding of what a crash really is. It doesn't necessarily mean that things get written down to 0 or some arbitrary level because everything has a price at which someone will buy.
Even companies have some value after a crash and you could make a case that at some arbitrary point it was worth $x and since the crash didn't cause the company to crater to below $x it has not "crashed". Even companies filing for bankruptcy have some residual value above what they might have been founded on - it doesn't mean the company hasn't gone bankrupt.
Yeah, I certainly did not mean go to zero, or near that.
To be more specific, I have often seen people argue for including crypto in a portfolio based on the theory that if equities drop a lot (25, 30%?) crypto will hold or go up. People make the same argument for gold.
I personally think Bitcoin will eventually peter out, and that it's fairly risky as it has definitely been pushed by vibes and promotion and everyone wanting to get in on it
I mean that can apply to anything. There's nothing intrinsic about gold that makes it valuable other than it's rare, but there's plenty of things that are relatively rare that aren't valuable. Yes there's industrial uses of gold but that's not why we as a society and a species treat it as valuable
Maybe bitcoin is the new gold and will hold value forever, and as more serious people get into it, it will lose its volatility and be less subject to the vibes shifting and there being a run on the market. Maybe not
It is different from Elong stocks, though, because the myth of Satoshi aside, it's not exactly a cult of personality like his companies are
To me the difference is that BTC's continued value isn't predicated on the actions of one person, who has continually proven himself to be childish, mercurial, and dishonest.
Yes when will everyone wake up for the charade of having a monopoly on space launches? Of putting over 90% of all mass into space, of your closest competitor being the nation state of China, and they are years away from where you are right now. Ah yes, that charade, when will people learn am I right? Total genius.
An extreme majority of Space X's self-determined value proposition has nothing to do with space. Out of an overall TAM of $28.5 trillion, $26.5 trillion is based on AI. Personally I believe that number will actually approach zero, but it certainly even in my most optimistic view is a number with a B, not a T
The total addressable market even at Space X's own calculation for space launches is only $370 billion. And, supposedly as the only company that can launch things into space they are still losing money on that business. This is bankrupt-a-casino levels of incompetence
SpaceX are currently losing money on the "going to space" part of their business.
They're making money on telecoms, and may have just started making a profit on renting out the data centres they originally built for the AI that it turned out hardly anyone wanted to actually use.
Michael Burry, a hedge fund investor featured in the book “The Big Short” for his predictions on the 2008 financial crisis, said in a Substack discussion last month that any increase in SpaceX’s stock after its I.P.O. would “be on hype and technicals.”
Here, “technicals” means technical analysis signals rather than the company’s business fundamentals. In other words, Burry is saying the stock could rise because of chart-based trading, momentum, and market behavior—not because investors think SpaceX is truly worth that much based on revenue, profits, or other fundamentals.
How long can the hype be maintained? TSLA is still maintaining its hype, judging by its P/E ratio.
The hype already starts with the SpaceX SEC filing. According to it, its addressable market is $28.5 trillion, of which $26.5 trillion are AI. This means that every human being who owns a computer on this planet (1.75 billion) would need to spend (on average) over $15.000 on xAI products.
>> that every human being who owns a computer on this planet (1.75 billion) would need to spend over $15.000 on xAI products.
Current US national debt is approximately $39.22 trillion. As we achieve a zombie movie, level of collective madness, lets take all this into the last degree.
Nationalize SpaceX, and with this bright obvious future described on the prospect lets pay US debt :-) Pay US retirement benefits pensions in token credits, pensioners can resell, make an options market for tokens. I am sure Robinhood will open you a margin account for that?
Lets open a futures markets for food goods grown up on SpaceX Asteroids, as they will have free solar energy. They can grow them three times faster than on Earth...
To quote Ron Baron yesterday on CNBC, we are all going to earn hundreds of billions....
While this would be conceivable if some future AI gets good enough to actually replace 100% of global paid labour currently done by using a computer, the reference class I have here is that Wikipedia is definitely not valued at [number of people online] * [peak cost of Encyclopaedia Britannica].
Economic displacement on that scale breaks the valuation.
It’s not our responsibility to provide counter arguments. You expressed that the two fundamentally different things are the same “in essence”. It’s your responsibility to define the two and explain how they are the same first.
Tesla definitely is floating down slowly worldwide with hype when it finally crashes just don’t be left holding the bag. 2026 will be another year downward.
Notably, it looks like the meme power is gradually being redirected from BTC and cryptocurrency in general towards AI and SpaceX in general. Now that people have found a means of consuming vast amounts of computing power that occasionally emits useful output, rather than just a hash and a colossal waste.
> Here's a shower thought. BTC essential is worth $70k solely through the power of memes. Can TSLA and SPCX remain overvalued (relative to the revenue of their respective underlying assets) forever through the power of memes?
The "value" of something can be a bit of a meta-game:
> A Keynesian beauty contest is a metaphorical beauty contest in which judges are rewarded for selecting the most popular choices among all judges, rather than those they may personally find the most attractive. This idea is often applied in financial markets, whereby investors could profit more by buying whichever stocks they think other investors will buy, rather than the stocks that have fundamentally the best value, because when other people buy a stock, they bid up the price, allowing an earlier investor to cash out with a profit, regardless of whether the price increases are supported by its fundamentals and theoretical arguments.
Plenty of folks may think these companies are garbage but are 'playing along' because it's not necessarily what they themselves think that is important, but what others think.
This idea was put forward by Keynes in his General Thoery publish in 1936, so human nature has hardly changed since then.
I'm not sure if this means gold isn't powered by memes or whether it's just one of the most long lived memes of all time. Aside from other nice properties like lasting a long time, being pretty, and not requiring electricity to exist.
Certainly, gold is not a meme, nor are all the other rare earth metals. This principle applies most to the other elements found on the periodic table. In fact, they have become more valuable due to research and development in today’s modern world, as numerous useful discoveries are made daily.
Cryptocurrency, such as Bitcoin, is undoubtedly a meme. If given a choice, one would prefer to possess any of the rare metals, rare earths, or any other elements that can be obtained in sufficient quantities, whether on land or in the ocean. Bitcoin or cryptocurrency is beneficial only to insiders and not to the general public. In essence, they are akin to modern tulips with a cherry on top, and like Tesla in 2030, one should avoid being caught with the bag.
What you are describing applies to all forms of money. It has value because people believe other people will use it as money. If that belief drops, the value drops.
People comment on gold and Bitcoin, but don't realize the same principles apply to US dollars and bonds.
Currencies are a little different, since they are required to pay taxes; and payment of taxes is enforced (to varying degrees) by state violence.
Hence if you believe you will be taxed ("death and taxes" being the only certainties in life, etc.) then the currency associated with that tax has value, in that it avoids imprisonment, etc.
I was once told bitcoin was a 'pet rock'. Thing is people pay a lot of money for rocks when they have no planned industrial activity for them. Diamonds, for instance.
I think you are spot on. The problem comes if SpaceX goes out of fashion, not its fundamentals.
This is why I think Elon buying Twitter was one of the most strategic decisions he ever made: he quite literally bought the meme machine that upholds the valuations of his companies.
I still don't think it was strategic on his part, he did try to back out after all; my guess is that having bought it, he tried to maximise the value it produced.
And given the financial statements in the SpaceX IPO, to the extent X still has any value at all, it is almost all just influence of one kind or another, not actual money.
At least as long as Musk is the CEO, perhaps. I don't think it's easy to find another charismatic figure to keep it going.
It's a different kind of hype than Nvidia has, which is showing very high and fast growing revenues (which may not continue, but they're real now). Jensen I think is not as critical to the AI hype as Elon is to his companies.
All these major tech companies eventually get leadership changes. Apple, Google, Amazon, have all done well because they're real companies and go beyond their original leadership. Tesla and SpaceX I think would surely go down the moment Musk is no longer in leadership.
Indeed, but for general stock market purposes it's one dollar one vote, not one person one vote.
Possibly even worse. Short-sellers got burned on Tesla (and perhaps now also SpaceX), which may mean the marginal buyer and seller consists only of people who already buy into the hyper and by trading are sharing price info with each other rather than with a single person who doubts the man.
I sold my Tesla stocks a while back; the people who kept them don't care what people like me think.
Well I meant he has a big Twitter following and whatever he gets into many are willing to fund and so on. Like there's no way Tesla would be taken seriously for all its future ambitions and valued accordingly without Musk at the helm.
I just cancelled my cursor, which was the best for my workflow. Will need to find something else.
On what grounds do you believe the value of BTC is meme-driven? Another (and arguably more plausible) explanation is that the price reflects the vast amount of criminally-obtained wealth stored in it. It’s a far better store than burying cash in mountain caches.
All money is meme driven. Money is fundamentally a meme itself. Seeing lots of people who seem to be making some sort of implicit distinction between bitcoin and USD and so on, but they are no different. They serve the function they serve because of what people believe about them, like any other social, cultural, or economic abstraction. Bitcoin has the feature of a verifiable ledger, but its value and function are in our heads, just like the USD or GBP.
That ship sailed long ago, before there were viable alternatives to BTC. There are plenty of ways to launder BTC (i.e. hiding in plain sight) and the existence of hoarded wealth in BTC isn't necessarily exclusive of other cryptocurrencies that are better fit for purpose.
BTC has the layer 2 lighting network that can process more transactions than Visa and MC combined with very cheap fees. I guess HN knowledge of Bitcoin does not go further than 2015
Also, everything on level 2 needs to be ultimately balanced on level 1; BTC doesn't have enough throughput to balance all the banks in the world, and lo Lightning appears to be consolidating into fewer nodes.
Also also, everything level 2 (including Lightning) necessarily takes away at least one of the selling points used for BTC, and replaces them with something functionally equivalent to 50% of a bank but worse.
> Current actual throughput is significantly lower, estimated at around 300 TPS, because the network is still growing. However, Lightning's architecture scales horizontally: as more nodes and channels are added, capacity increases without any protocol changes. The network's total value locked has grown past $500 million, with over 15,000 active nodes and 50,000+ payment channels.
The layer 2 has been predicted by one of the first adopter in 2010 so it was always the plan
SpaceX the company has done and is doing amazing feats of technology.
SpaceX the stock is only in a very small part about rockets and space. In their filing they talk about their total addressable market being $28.5 trillion, of which $26.5 trillion are AI.
That means that they believe the SpaceX stock is 93% about AI.
It's open to argument whether xAI or BTC is crazier - I just wanted to say that when talking about SpaceX stock, it's not really about rockets.
neither of them are sane. BTC is useless, unless your trying to buy child porn, buy illicit drugs on the internet, or someone who bought it before the value exploded. eventually, the world will come around and it will go to zero, if quantum doesnt kill it first. im looking forward to that day.
Sure, if there were a stable enough market in which I can fetch twice that amount for that tulip, I'd buy it in a heartbeat. Wouldn't you?
The point being: BTC is a an abstract good, of no practical use except that of being transacted. Has whatever value the people are willing to pay for it, and has had a value in the tens of thousands for long enough that buying one with the intent of keeping it for a while is not such a stupid idea. I don't currently own any but there is a price at which I would buy one, and that price is many thousands of dollars... For an alphanumeric code in a distributed ledger.
Being transacted is the least usable thing for BTC. He you tried to transact on it? Its so slow its funny people claim it a use case for it... I know of only one use case, and its to anonymously bribe politicians who have their own coins.
True, it's very useful for scammers, grifters, international terrorists and authoritarian governments to funnel monetary value around without having to rely on the traditional financial system where they may be blocked and their money confiscated.
It costs 150 usd to send 5k usd between two countries I live in.
It costs the same amount of money to literally go there by plane and bring it as cash. This is not fair obviously. So traditional finance is a scam itself.
It takes 10 seconds and no fees to make the same transfer via blockchain.
I am pretty sure you know basically nothing about those crimes or the people that do crimes or how they actually transfer money. Just doing casual newspaper intellectualism while talking about things you never interacted with
> funnel monetary value around without having to rely on the traditional financial system where they may be blocked and their money confiscated.
Yes. Unfortunately the traditional financial system is governed by a country which has been behaving increasingly erratically, threatening its long-term allies with invasions and committing obvious war crimes. This is not a "good guys vs bad guys" scenario.
As long as there is someone around who is very good at keeping the price inflated (and that in turn also because he did actually deliver extraordinary things, it's not just smoke and mirrors).
On the other hand, the fact that BTC has absolutely no intrinsic value can be an advantage over a real company, as it makes it more insulated from reality. Supply chain shock? No problem. Competition? Same. New technologies, political change? Neither.
SPCX only floated a small amount of shares, and made the stock exchanges compete so that it would get to rig the system in a way. If funds have to buy SPCX + small share amount, it's going up. This is the reason stock buyback used to be illegal. It's purely market manipulation of share count, not market based on actual value.
> I saw that one wealthy individual had purchased $1 billion of SpaceX at the IPO. Does that count as a retail investor?
There were two individuals who each bought $1B: Ron Baron and Gina Rinehart.
While they are individuals, they executed these billion-dollar investments through their massive corporate entities and investment firms, rather than personal brokerage accounts.
A retain investor is an individual, non-professional investor who buys and sells securities through brokerages using personal funds.
At least some of TSLA and SPCX value is derived from Musk's ability to purchase politicians to ensure the tax dollars keep flowing to them. Essentially, these ticker symbols are backed by the US government's ability to force us to give Musk's companies our money.
Mostly because we are looping AI to fix problems, and then the same data is used to improve AI. There is no upper limit to this.
Taken to its logical conclusion, this process needs a hardware scale that might even look laughably huge at this point. Its fairly obvious space is going to play a big role in the coming times.
I could be wrong, and I humbly accept it when Im proven wrong. But it does feel like a lot of people in top places know we are going to need all the energy and resources space has to offer to run this runaway intelligence.
> we are at runaway intelligence already. Mostly because we are looping AI to fix problems, and then the same data is used to improve AI. There is no upper limit to this.
I'm an AI booster, but this really doesn't follow. There's absolutely no guarantee that the marginal improvement from this continues to hold.
Validated data coming out of LLM(Itself generated from a recursive/loop process, used to incrementally arrive at solutions) being using to improve the very LLM is a very powerful loop. And there is no real upper limit to this, at least not in the near future.
Like most exponential processes, the start is slow, but it gets fast very rapidly.
This is also what Anthropic has said, the ones training these models today(i.e 2025 - 2027) will be impossible to catch up with, let alone beat.
It seems like that during a bubble (things can never go down!) but the market eventually does correct, even if it can take years or even decades to do so, and usually overcorrects when it does.
It's down to the balance between buyers and sellers. If you've got more selling than buying it'll go down but Musk has been remarkably good at keeping the buyers there.
Don't forget the amount of shares released to the market. That impacts the price dramatically as can be seen be all the 'stock buyback' manipulation of stock prices.
Indeed. There's bit of a funny effect just now because the various index funds are having to buy into SpaceX so their funds match the index and apparently that's ~$14bn of forced buying at a time when there are not so many sellers.
Someone invents something that is digital, but can't be copied. Quite brilliant as it is the first digital asset that can store value without centralization and trust, based on market demand.
Someone on HN: "BTC is valueable solely through the power of memes".
Bitcoin doesn't "store value", it has the value that society assigns it, which is what the parent means when s/he says "BTC is valuable solely through the power of memes". It's not a fiat currency, nor does it have any intrinsic utilitarian value.
> Age old argument: money also has no intrinsic utilitarian value.
I didn't claim money had intrinsic value, I claimed that money is fiat currency--it's backed by a lot of people with a lot of guns. BTC is not.
> And yes, btc does store value, it is doing that for me now. I stored some of my value in it and it has held better value than fiat.
Currencies (BTC included) don't "store value", they are assigned value by society esp via markets. Sometimes society decides one currency is more valuable than another, or that its change in value differs from that of another, but neither currency stores value.
The overwhelming majority of SpaceX holders are institutional. They had planned to allocate 30% to retail, but it ended up in the 20% range as a result of institutional demand. [1] No clue what's going on right now as their stock is going to the Moon. But in any case, I think the people that don't understand why it's doing well are mostly those who are unfamiliar with the space industry. SpaceX has already revolutionized space by dropping the cost to orbit by multiple orders of magnitude relative to the Space Shuttle, which it replaced. And it looks set to do that again with the Starship.
The main reason humanity hasn't meaningfully started expanding into space is because it used to cost $54,000 to get a liter of water into space. SpaceX brought that down to about $5000, and then further down to near $1400. That's a massive reduction in price, but you're still left with the problem that it costs $1400 to get a liter of water to space, which is why we still can't have nice things, yet.
Starship has the promise of bringing that down a couple of more orders of magnitude where the goal is to get it within the $10-$20 range. If they succeed, then you've just opened the doors to an entire new frontier of expansion and growth for humanity which is practically infinite. And right now there's no real reason to think that they won't succeed. And more importantly than this is that nobody seems to be able to compete on their level, or even remotely close. Their closest competitor is probably China who remains technologically well behind. And so SpaceX today is akin to being able to get a piece of some sort of super-ship making monopoly, just prior to the Age of Sail. The downside risk is basically zero since they're still making rapid progress - the only question is how rapid. And upside potential is basically infinite.
For why we'd want to go at all: there's a lot of resources up there, and pollution is much less of a concern for factories made up there. Also some material processes may be much easier in zero-gee.
But that doesn't mean it's actually worth the effort.
Asteroid mining is the obvious one. That said as part of looking into SpaceX's business I learned that currently the TAM for launching other people's stuff into space is under $10 billion. SpaceX already owns most of that market. Their own focus on data centers in space IMO speaks for itself.
to be fair, in the pharoahs quest for immortality after death, if they could have put their pyramid burial tombs in space to avoid decay/ruin for longer they would have. I wouldn't be shocked to see the oligarchs paying for their own space burials inside very lavish satellites whose orbit won't decay for thousands of years so they can "live forever".
While I am *extremely* skeptical, I've heard people talk about space-based solar. I mention this only so nobody adds an "and also" for that.
Ditto anyone talking about lunar He-3. This is a bad idea, "why" is long enough to be worthy of a blog post.
But if you were committed to expansionism for whatever reason, then you'd want some way to get industrial quantities of steel from mines already in space, to make the factories themselves cheaper. Possibly megastructures like an orbital ring, too.
The hand waving is probably reasonable for a couple of reasons. The first is that you can't really math things out yet when costs are shifting by orders of magnitude on relatively short timeframes. What seems impossible in one year is a casual achievement in the next. If we go back to 2002 (when SpaceX was founded) and I told you that within 2 decades they'd have lowered the cost to get stuff to space by more than 97% relative to Space Shuttle, it'd have sounded somewhere between unreasonable and impossible, yet that is exactly what happened. And Starship is set to send prices plummeting yet again.
The second reason is that the notion of economics becomes kind of odd when you start introducing space. Space has practically infinite resources of every type imaginable. For instance one single asteroid, Psyche 16, is thought to have orders of magnitude more gold than has ever been extracted in human history. But I mean, what does this even mean economically? Obviously if you start bringing back hundreds of thousands of tons of gold, the price is going plummet. And even the viable possibility of this happening would probably send the price of gold plummeting. So... it's weird. We're in uncharted waters.
And then on top of this access to space will mean we will start expanding outward - colonies, vastly larger stations stations, and so on. And these expansions will develop their own parallel economies with their own needs/production/etc. This is all kind of like trying to predict the economics of the internet before the first cable had been laid.
Most people I see online freak out at the idea of normalizing living in an apartment/condo and slightly higher density. I don’t see them lining up to live in space :)
As far back as the 60s NASA had already laid out technical plans for how to get humans to Mars. The Moon was never meant to be the destination, but a brief stepping stone to greater things. So we can thank Nixon for setting humanity's progress back by half a century. The ISS and Space Shuttle both were driven largely by these initial plans.
But the current claustrophobic nature of the ISS was, again, not the destination. There was a grand scheme of in-orbit refueling, fuel depots, and much larger scale space architecture. Something of the ISS scale was intended as a short jumping-off point, a workers' tent analog. But our progress stalled out and the workers' tent is what we were left with.
The point I'm making here is that space in the era we're stepping into is very unlikely to look much like the one we're leaving. This is even more true as the price drops because private industry will be able to play a major role. There will be no Nixon to just cancel everything out of concerns for his political career.
Depends on the timescale, and what other tech is the prerequisite to this and what other things can be done with that tech.
Even with cheap rockets, Musk was talking about 6-digit-dollars for becoming a Martian, I don't see more than a few million people wanting to spend that much for that outcome.
If we get self-replicating robots or whatever, big space stations aren't that expensive, but also Earth is huge and you can extract minerals from seawater electrolytically and PV is already cheap, so perhaps we'd just get a load of seasteads instead: https://en.wikipedia.org/wiki/Seasteading
Really hard to say. Everyone who does space stuff talks with the same level of optimism as Musk but historically they deliver less.
But for what it's worth, I don't think Musk's going to reach Mars except in an urn, baring some surprise development in anti-aging medicine. Even though their progress is rapid by the standards of the industry, they are still slow in terms of natural human lifespan and his age.
> Have you sketched out the net present value calculation for an investment in spacex today based on that?
IMO, their market cap should be around $200bn at the moment. That's mostly Starlink, so if there's a Kessler cascade *or* if China makes a competitor that goes down to about $80-100bn
The long-term stuff about Mars or turning the moon into a factory for modular data centres is too far into the future to be worth considering.
That was my general impression. They did amazing work to totally change launch costs (with more on the way). But launch was a limited market. The low costs have and will grow it, and starlink will grow it, but not enough to justify this value.
In investing it's critical to think about things probabilistically, rather than trying to just guess the future. "Fooled by Randomness" is a great book on this. In a scenario like this it's especially critical, because in the case where Starship fully delivers and the next great frontier of humanity opens up, the upside potentially is practically infinite.
E.g. imagine in a bet you think there's a 99% chance of something being worth $1 and a 1% chance of it being worth $1 million. What's a fair price? It's a simple calculation - 0.99 * 1 + 0.01 * 1,000,000 = ~$10,000. That's you setting a price of $10k on something that you fully expect to be worth $1 99% of the time, and still coming out just fine.
So the actual debate should not be on 'guess the future' but rather on the odds of Starship delivering and estimating the impact this would have. And I think to many the answers are fairly easy - 'very nonzero' and 'very big.' It's akin to the NVidia/Apple stories, except in this case it's being somewhat priced into the market to begin with because it's somewhat more probable, and easier to foresee.
The problem with that argument is the failure to reasonably factor in time. If you take an expected value approach (quite reasonable in general) looking into the future you need to discount for time or everything goes to infinity.
While the point is valid, the upside still isn't practically infinite; humans intuitively do something close to hyperbolic discounting for future returns (I assume evolution did this because of the possibility of black swans but it isn't talking).
The easy version of this is to value a company over the next 20 years of profits. The first half of that time horizon, I don't see much changing at all outside of Starlink subscribers. They've only got ~5 Mars launch windows in any given decade, and while their design philosophy is great for LEO where launch windows are daily, launching rockets to see what fails is exactly wrong for Mars.
OTOH, now to 2036, there's at least a chance AI goes boom. Perhaps even one of Musk's AI. Great for the overall economy (at least right up until it changes the economy so much that capitalism looks as obsolete as feudalism looks today, which may or may not be in that horizon), but even if it's the next industrial revolution, there's so much competition there that I expect everyone's profit margins to be almost nothing. And conversely, if any one of them does appear to be massively dominant, in comes an anti-trust case (who remembers Standard Oil?)
Same with rockets, though a "rocket boom" is perhaps an unfortunate phrasing. He finds a way to make more money than expected, just like he's already doing with Starlink? Great. Industrial espionage is a thing, and even if it wasn't, China's not got a problem with the state funding businesses to simply re-discover everyone else's secret sauce. US companies and residents may find themselves restricted from using the fruit of that work, but most of the world will just buy what's cheapest.
Starship has already made it to orbit multiple times. It's likely that it will be ready to go operational in well under a decade. As we're speaking of upside potential, let's just assume the aspirational goals are hit - and you end up seeing $10-$20/kg to space. You've just revolutionized not only space, but all transport on Earth as those prices are less than most terrestrial shipping options, and it can be delivered anywhere on Earth in a matter of minutes.
The implications of that are quite difficult to even fathom because again this is one of those things where I feel like we're trying to predict the economics of the internet before the first cable had been laid. But the one thing that can be very safely said is that the impact would be revolutionary.
And that is completely ignoring space when the whole point of Starship is being able to get 100+ tons to the Moon or Mars, which again revolutionizes those domains first in terms of size, but vastly more importantly - in terms of cost as well. This opens the door to viable private development on the Moon and Mars (to say nothing of space itself). And again the implications of this are very difficult to predict, but it would also certainly be revolutionary.
---
The competition issue you've brought is, in my opinion, one of the most viable concerns. If America can get stuff to orbit for $20/kg then it feels like China always finds a way to do it for $0.20/kg. But the reality is that SpaceX's secret sauce isn't all that secret. Their exact alloys, general construction specs, and more are all public knowledge. Many have tried, including with vast resources and wits behind them, and all have failed.
China is almost certainly going to be the #2 player and their Long March 9 [1] is their attempt to clone the Starship, but for now it's vaporware. And to actually get these things out onto the launch pad isn't the endgame, but just the beginning of the race. They are almost certainly years behind SpaceX, and the distance is likely increasing over time as SpaceX shows no signs of slowing down, and now has no funding constraints whatsoever.
> Starship has already made it to orbit multiple times. It's likely that it will be ready to go operational in well under a decade. As we're speaking of upside potential, let's just assume the aspirational goals are hit - and you end up seeing $10-$20/kg to space. You've just revolutionized not only space, but all transport on Earth as those prices are less than most terrestrial shipping options, and it can be delivered anywhere on Earth in a matter of minutes.
Let's assume it reaches $10/kg (I don't think it will, but let's assume that).
A 20ft shipping container can carry about 24 tons; the cost of shipping it to the other side of the world is much less than $240,000; the numbers I see vary depending on details, but even high numbers are less than $6k, which is $0.25/kg, and even then half of it is port-side not the ship's own cost. Air freight may get more competition, but not all shipping. And it won't replace a lot of trucks because it's too big (how many places want 10 container loads worth of goods at the same time?)
Yes, point-to-point 45 minute deliveries worldwide are going to have their uses, but price alone isn't revolutionary in that particular way.
Passenger flight would be something I can't rule out, but multi-gee launches will limit the numbers, the inherent noise issues will limit the launch locations.
I do agree with you that the implications will be difficult to guess correctly; however this is true in both directions, not just the positives, e.g. network hardware is vital to the internet, Cisco makes that, yet here's the history of Cisco stocks, with 25 years from the peak of the dot.com bubble to when it next reached the same price: https://www.google.com/finance/beta/quote/CSCO:NASDAQ?window...
> And that is completely ignoring space when the whole point of Starship is being able to get 100+ tons to the Moon or Mars, which again revolutionizes those domains first in terms of size, but vastly more importantly - in terms of cost as well. This opens the door to viable private development on the Moon and Mars (to say nothing of space itself). And again the implications of this are very difficult to predict, but it would also certainly be revolutionary.
Eventually, yes. But not soon enough for me to want to price it in today. Remember, hyperbolic discounting, black swans, competition, other things I can do with the money while I wait, all that means when I price it I want to know what it can do in the next 20 years at most. Even if I'm being optimistic about timelines, it would still take 20 years for the Moon to get enough industry for SpaceX to be worth the current market cap. (Well, unless someone makes a von Neumann replicator, but if that happens then whoever makes the replicator wins, regardless of if it was launched by SpaceX or not). Mars is harder.
> China is almost certainly going to be the #2 player and their Long March 9 [1] is their attempt to clone the Starship, but for now it's vaporware
Until they have enough confidence to put it into a circular orbit rather than one which has it hitting the Indian Ocean every time just in case something fails, the same can be said of actual Starship.
> They are almost certainly years behind SpaceX, and the distance is likely increasing over time as SpaceX shows no signs of slowing down, and now has no funding constraints whatsoever.
I assume China's behind, but with their secrecy (let alone the language barrier) it's hard to tell.
But SpaceX definitely has funding constraints. GAAP is says about $60bn obligations, but all-in they want to spend several hundred billion over the next few years and do not have enough revenue to cover it, so even with the IPO they are at actual risk of running out of money.
This is why the IPO happened at all: they ran out of private investors willing to give them money.
Most of your arguments have been perfectly reasonable, but trying to argue about the value of a shipping tanker vs point to point minutes time transport is not. We're speaking of viable same day delivery anywhere in the world, or same hour in some cases. Obvious complete replacements would be air freight, military logistics, and much more. For stuff you're fine taking months, sea shipping will likely remain favorable.
Cisco is more comparable to something like NVidia in modern times in that their valuations was driven by success during a time that was already screaming bubble, and an expectation for that bubble to somehow never pop. By contrast the space industry hasn't even really begun yet, so the wager on SpaceX is more speculative in one way, but also much more likely to be stable in another.
China isn't likely behind - they are definitely far behind. They can't even settle on a design for their Starship. And similarly, I think one thing SpaceX has proven is that fail fast is the way forward. You can't just sit in a sim and send these things up. There's too many unknown unknowns that you can only learn by watching things go boom, a lot. There is no world where China just unveils their secret Starship clone and starts flying successful missions with it. Again that's something everybody tries (in no small part because it sucks when things worth tens of millions of dollars go boom), and always fails with.
SpaceX has never had any trouble raising funds when desired. I think the main reason they IPO'd is as a black swan hedge. We're at a major inflection point on many levels, and it's very conceivable that in 10 years the US' role in the global economy will have been greatly reduced. Obviously that $3 trillion valuation is largely funny money, but for now that funny money still functions well enough. But that's likely liminal. It's entirely possible, if not likely, that we're currently Japan circa 1989.
It's not about the "Access to resources. A sense of adventure. Learning."
Since you are right. All of these things are available on Earth for far cheaper, with better ROI and with far less risk of dying or getting early cancer from cosmic rays.
The real benefit is a commanding lead over geopolitical rivals in the domain of satellites and satellite platforms for hosting devices in orbit, thus securing an insurmountable national security advantage.
The only argument for space is because you're so set on capitalism that you'd rather destroy the earth than stop trying to continually grow profit. If you're that sick and deluded, then space gives you a way to escape the consequences of your own foolish hubris.
So, your argument for SpaceX is that they'll take physical systems that they've already tried to squeeze down, and squeeze them down nearly two orders of magnitude? What fundamental scientific discovery do you think is going to enable this? Or do you think that AI is magically going to do it for them?
That has nothing to do with how they have already reduced costs, and how they continue to do so. Before SpaceX, you launched a rocket to space once, and then you discarded it. Imagine how much a plane ticket would cost if each time a plane was flown once, it was then discarded.
The big revolution with SpaceX was meaningful reuse (I can get into the comparison with Space Shuttle if you're unfamiliar there). Landing and reusing rockets is something that Boeing et al thought was impossible from an economic point of view, and actually taunted SpaceX in their early years over it along the lines of 'Oh you're going for reuse. Yeah we researched and trialed that out a decade ago. The economics don't work. Cute to see you trying though.'
Their success there is what helped bring the costs way down. But there's still plenty that's not reused - in particular they currently only reuse the first stage (the big rocket that gets things off the ground initially) while discarding the second stage - the space-optimized payload delivery rocket. With Starship they're going for 100% rapid reuse. So you're looking at this absolutely massive 2 stage system where both parts will be able to be repeatedly and rapidly reused.
You really think cost is the main reason humanity hasn't expanded into space? I'm all for space exploration and learning from space but actual humans are quite squishy, like gravity, dislike radiation, and would need to take a lot of water with them just to visit a rock very indifferent to their existence.
Definitely. Humans expand. It's probably an evolutionary imperative in many of us. Africa was essentially an oasis and yet some of us decided to go make our way outward to inhospitable freezing ass places where the environment itself is just constantly trying to kill you in a million different ways. And we knew nothing about how to deal with it, at first.
Or take the early sea voyages. Not only were there endless worries about things that were ultimately nonissues like sea monsters or falling off the edges of the Earth, but there were endless very real dangers awaiting which we had no clue about or how to deal with - scurvy, rogue waves, and much more. In the aforementioned Age of Sail, it was just expected that a significant chunk of your crew would die. Yet somehow we pushed onward and outward.
And the infinite possibilities of space are going to absolutely dwarf all previous frontiers in terms of interest and potential.
The limits are when companies and institutions start to default on their loans. Or, in the case of governments, trigger hyperinflation by printing money to pay off the debt.
Of course it can collapse before that, but if it gets to that point it is guaranteed to collapse.
But that won't happen while the share price keeps going up no matter what. Borrow more, secured against the massive unissued equity, or issue more shares.
> BTC essential is worth $70k solely through the power of memes
Like everything else in finance...
Saying this is not to defend all sorts of crypto-bros. The economy, especially one overly focused on publicly traded companies like the Western, and especially the US economy, is a meme economy.
Coffee, flats, healthcare, military spending, etc., of comparable quality in the abstract East, cost multiples less than in the EU/USA because they and their currencies are weak on memes.
Cursor was my first hands-on experience with AI. I didn't know much about getting set up with specific providers via API, and Cursor made it easy to pick any model, ask a question about some code, and get a clear suggested answer easily viewable in the IDE with an 'accept' or 'reject' button. I think they answered this question well: "How do normal developers want to interact with AI?"
I moved away from Cursor when I noticed the responses from specific models were not as clean or accurate as when I'd prompt the models directly, which was something I didn't know how to do early on. I hypothesized that they had some boilerplate prompt sitting atop of my own, causing less precise or desirable results.
I would assume Cursor is still one of the best options for normal developers to get started with AI, but with Copilot forcing their foot in the door at many companies, I wasn't sure how well it would fare on its own. Being acquired by SpaceX should help, and I'll be interested to follow along and see how things develop.
My opinion is that the publicity can only help Cursor. I don't necessarily think SpaceX would make Cursor better. Copilot (which I view as a direct competitor to Cursor) has a huge structural advantage. I have several friends in various American companies where Microsoft products are all they are allowed to use. They get "free" Copilot access as a part of their Microsoft plans. Developers aren't having Cursor placed right in front of them in the same way Copilot is, and from my experience, when developers have the choice to pick one, they pick Cursor. So, I just feel the SpaceX/xAI publicity could help Cursor get more visibility in these general American software companies more so than they could on their own.
When I first saw the company built on top of vscode in such a crowded field way back at the end of 2022, I thought "forget having a moat, they are renting their castle from the invaders!" - I couldn't see how see how a single team could execute well enough to effectively muscle their way in between Microsoft and OpenAI, who at that point looked destined to control the developer ecosystem between GitHub, VsCode and the then-best coding models. I think it's easy to forget how insane this seemed even just a few years ago.
But every year since then they managed to simply ship a better product on the axis that mattered to the most users. And now they are sitting between a huge user base and a massive stream of valuable tokens, they can sell to SpaceX. Incredibly impressive.
I am out of the loop with Cursor. Did they make their own LLM model or do you mean they just capture what is being sent to other models and that is valuable?
They have a few in models that powered the app, including one that applied edits. Recently they also started fine tuning Kimi models under the “composer” brand, and Composer 2.5 is a very cost effective coding model. But I suspect that the real value is in the distribution they have, which is what I primarily meant.
Happy pi.dev user here, give it a try! I would say that's kind of the "vim experience" but for harnesses: has the minimum, if you want something more you extend it :)
Saying it gives "the minimum" is generous, it's pretty much useless out of the box. And did I say slow as well? I think pi is great if you're into spending your time managing your harness rather than using it. In that regard, it's more like neovim.
The idea is to make it fully autonomous so it is not really something that is meant to be constantly prompted and it is unlikely to fit most workflows but the idea is to make something that fits the future - not the present.
I am seriously thinking about adapting Kate with a fully opensource ai harness. it should be good enough for mac and linux for most devs. it already supports lsp server & has a established plugin infra so it should not really be any blockers. anybody wanna collab?
this class of spyware pretending to be ide makes me sick.
Nvim + Claude code. Or zed. Honestly basically every ide is adding agent harness features as their primary focus right now. Just throw a rock and you’ll hit something that will work for you. You can even just use emacs and have ai build harness features for you.
"Back in the day" is the key idea here. It's the valuation that doesn't make sense in today's market. What does Cursor offer to be valued at $60 BILLION? Its models are alright but nowhere near SOTA, and pretty much superseded by open source local models at this point.
Composer 2.5 worked just as well for me as opus 4.6, and was faster and actually worked better for quick bugfixes and reporting back to support. And is much cheaper.
I'm not saying HN should be super supportive of everything, but the level of hate and complete loss of reality for a lot of people is quite sad to see, for a community of supposedly intelligent people.
I don't know if it's just me, but the thoughtfulness of of the average HN comment has seemed to decrease in the last few years. Feels more populist/Reddity the last couple years.
You haven't been paying close attention. This has been going on for about 15 years. I'd say we had 5 good years here before the quality began dropping precipitously. If you got here after about 2012, you're probably part of the problem.
I'm pretty sure it has. If you look at threads from a decade ago, they're completely different - much less dramatic, much higher quality, way less populist and full of manipulative language.
Nothing wrong with Cursor but $60B, wow. How many of these deals in 2025, 2026 will be worth nothing in 5 years? Seems like everything is just desperation and less like long term strategy.
Wow everyone on HN seems very out of touch with Cursor.
Their UX on their agents only window is quite good and on par if not better than Codex and Claude desktop apps and it still has quite a good bit of subsidization especially with Composer 2.5 being on par with Sonnet at the very least. They're also growing tremendously this year to 4B ARR recently on track to go higher before EOY.
Exactly. People here still think that Cursor is just a vscode fork while it has positioned itself as agent orchestration console that happens to have an editor you can switch to.
I don’t really understand what the value prop of cursor is, it must be the data and models. These days programmable editors like neovim and emacs have a huge advantage. I’ve had ai create several custom plugins to have my editor do whatever I can think of. Just ask Claude code, hey I want to do x, y, z, it spits out some lua and I have a new capability. I don’t know why anyone would want to be limited by an extension interface at this point.
But they (SpaceX) could have backed out of the deal at any given time as they had the option to (and be required to pay the 10B break up fee). Nobody knew what would happen at the time.
This announcement is a definitive agreement of the acquisition at that $60B valuation.
I don't think you can even say that without reading the actual announcement. [0]
Was the acquisition "effectively announced months ago"? or was it the right to acquire the company or pay $10 billion that was announced months ago?
It would be equally relevant if it went the other way. But clearly both of you are confused on what was said they would do in the future (which that was the announcement months ago) vs what are doing today.
They're getting paid in extremely overvalued stock, so maybe it balances out.
This is not really a diss on SpaceX either because a lot of IPOs go through an immediate pop and then 1-2 years of doldrums as lockouts expire and promises aren't quite delivered.
Nobody knows what 60 billion in SpaceX stock today will be worth when Cursor insiders finally get to sell (at least a year from now, after other SpaceX insiders have started selling).
Generally this is how liquidity works. Their employees will have a six or twelve month lock up (six being most common).
Investors in certrtain rounds (or sizes) tend to have no lockup, whereas later stages have a six month. Alternatively, I've reviewed agreements where the lockup is based on minimum market cap, but I've only seen that a couple of times.
Ok everyone saying this is how it works but where’s the proof? SpaceX has a 7 day lockup for some people which is abnormal. So clearly the way it’s done isn’t in fact how it’s always done.
This is just a general practice that always happens when paying in stock. It's to prevent a massive dump the next day which would tank the share price 'artificially'. Again, rich people's rules.
Ok everyone saying this is how it works but where’s the proof? SpaceX has a 7 day lockup for some people which is abnormal. So clearly the way it’s done isn’t in fact how it’s always done.
> Holding Period. Before you may sell any restricted securities in the marketplace, you must hold them for a certain period of time. If the company that issued the securities is a “reporting company” in that it is subject to the reporting requirements of the Securities Exchange Act of 1934, then you must hold the securities for at least six months
Almost guaranteed to be. The only other way would be to have issued them during the transaction with the Sec. There was no mention of that. 60billion of issued shares would have been mentioned. It's non trivial amount.
Um, I honestly don't know for sure, if I'm not mistaken SpaceX has some weird staggered release schedule for employees and early investors, I guess based on their stock dates.
> SpaceX has a 7 day lockup for some people which is abnormal
You might be referring to staff members who have shares ? Their shares are not restricted securities as far as I know, but their internal company policy might affect those, but I'm not 100% certain on that.
There are many tranches here. Some friends and family got to buy day one IPO with no restrictions. Then some employees get a rolling release starting June 30.
Ok everyone saying this is how it works but where’s the proof? SpaceX has a 7 day lockup for some people which is abnormal. So clearly the way it’s done isn’t in fact how it’s always done.
Well, the people who bought the SpaceX IPO essentially footed the entire bill here. And they might still make money on it, depending on how the stock goes from here on. I don't see anyone who could lose here, even if the bubble bursts, apart from the Cursor people. And they are likely still going to make a huge amount of money.
> Well, the people who bought the SpaceX IPO essentially footed the entire bill here.
It's hard to say that they footed the bill here, but they basically gave SpaceX a number to say "well our stock went IPO and it's at this price, so here's 60B at this price"
A good tactic from SpaceX as after the inital surge of a big IPO, the stock price usually comes down and finds it's correct balance, which is usually always lower. So if they had of waited the 'cooling off' period of a year for example, and the stock price went down to it's 'correct' valuation, then they would have had to issue a higher number of stocks.. At least that's my thinking, but I'm terrible with money.
No, look a Composoer 2, it stands out starkly on its own in the pareto frontier on low cast and fast models.
Composer 2.5 was a huge leap with minimal compute from xAI.
They can compete with OpenAI and anthropic with xAI scale compute. They have a top notch model team and incredible training data and huge enterprise costumer contracts.
my employer (one of those huge contracts) dropped cursor in favor of claude and i don’t think this is true at all
while we had it i used cursor for probably eight months as my main ide (i did really like the interface for embedding code in prompts!) but had no problems switching to claude code. i asked around, and i truly don’t know a single coworker who misses cursor even a little bit.
I was fully in on Cursor for a good chunk of last year, using Composer + Gemini Pro (via Copilot / GH integration). I really enjoyed Cursor's tab completion capabilities, but when Sonnet and Opus started getting particularly good for me (think for me it was around 4.5), I swapped over to Zed + claude code in the integrated terminal. I've found that after a bit, I haven't ended up missing the tab completion. I've been perfectly fine with just LSP + claude always open. I don't miss Cursor. All my colleagues are on claude code with half of us also using Zed.
For someone who is new to agentic code or is generally somewhat junior, Cursor is very easy to get started with and is generally fairly frustration-free.
I use a cheap $20 subscription mostly for occasional use of Opus and Composer.
SpaceX made a smart move here. Someone else should have really seen the opportunity and bought them.
How the hell is an IDE a "pareto frontier"? Even if, say composer 2.5 is a huge leap forward, that doesn't mean IntelliJ or Vim or Emacs or Codex got worse.
Nope. Vim and Neovim users can use the cusor-agent cli for agentic stuff and there prefered editing tool for editing. All the major providers have a cli specific version these days. Probably because folks actually didn't want to actually use the cursor Gui and once Claude code came along those folks jumped ship and went full cli again
This doesn't match the empirical observations of a lot of people I'd trust more than you, and putting it below Haiku immediately makes it extra sus (Sonnet would have been the credibility preserving comparison).
Hmmm. Not in my experience. I don’t think it can be compared to Haiku, maybe sonnet levels? It’s obviously not Opus and never was intended for that use case. I use it quite a bit and it works well and is extremely fast for the tasks it was built for
It doesn’t tell anything about the valuation, but I prefer it over Claude Code, and I even stopped using JetBrains IDEs because of it.
Vs Claude Code: I like the option to change the models, as I often prefer ChatGPT or Composer to Opus. I have a slight preference towards TUI, but not so strong to drop the models.
Vs JetBrains. I really love JetBrains but the tab complete just works so well for me.
its not just the models, their auto complete is actually really good. when you make a change it will give you "tab to next" which makes refactors super easy.
composer 2.5 is also a very decent model, it go 90% of my AI tasks using it now.
I'm not sure it's worth the price tag for developers though, I mean resharper promised similar, delivered half, bought by millions and still don't think Jetbrains, even with its other good tool suit is valued 60B.
Yes, that's why Cursor was very popular when actually reading model output paragraph by paragraph was still the way you used them. That's no longer the case, their use has cratered, and in fact they have been disintermediated by their model vendors, leaving an empty shell.
composer 2.5 is literally just a fine-tuned kimi model, and the autocomplete is exteremly meh.
The only kind of AI I want in my editor is an autocomplete, but this isn't very magical to non-programmers (their TAM) or all that valuable (you can't charge thousands), they bought Supermaven and basically killed it, I'm not sure how you think tab is really good, I've not been impressed when I played around with it.
I'm honestly not up to date on all the FIM autocomplete at the moment. However, autocomplete getting good is something I look forward too. I believe its mostly a problem of latency. You needs <200ms (and accurate) for it to not feel annoying, which is hard to do and use a larger more capable model.
Unfortunately it doesn't seem like this is a problem companies are looking to tackle because it doesn't really make you any money or impress investors. Inceptionlabs.ai seems to be making attempts, but I haven't tried. They are apparently in the Zed editor, something else I haven't used.
Cursor was nice when I was still meticulously hand coding my stack, fantastic autocomplete. With today's top models, I barely write code myself, just review commits. Cursor eats Opus credits like there is no tomorrow. Composer has been a net negative in my experience. All in on Codex with GPT 5.5 on high using /fast.
Cursor, from my companies perspective at least, seems to be handling charming leadership to get enterprise AI contracts in place, compared to the alternatives. That's feeling like the moat from my first-hand experience. Easy single contract that covers a lot of AI cases that management wants to say they have in place.
Could not say on the valuation front. The one thing that is obvious is the trash talkers love to point out how horrible it is but I have yet to find a compelling replacement. Obviously Claude Code works but at least for me I never got along with the CLI workflow very well and sure I could use vs code with the extension for Claude but then I lose tab autocompletes which I actually like. I have yet to see anyone build a derivative model like composer 2 that is quick, cheap and has higher reliability on tool call use. Again I don’t know on valuation but it’s pretty impressive how far they have come. I look at Jetbrains and at least from an AI perspective they have been left in the dust.
Using Cursor with Jetbrains via ACP isn't that different than using within the Cursor IDE. You still get Agent/Plan/Ask modes with model selectors.
I'd recommend installing the Cursor IDE, then using it to install Cursor CLI (it's easier to keep things up to date this way), then setting up your Jetbrains IDE to point to the Cursor CLI via ACP integration.
Cursor's moat is that it is a virus that infects organizations through shared skills, hooks, agents, etc.. Once one person uses it and infects the repo everyone else starts using it.
Know a few companies that have moved from it fully to Claude. It’s still early so the moving cost is low and Claude Cowork is something non-tech employees can make use of much easier than Cursor. I really don’t see what Curor’s value is longer term. Why pay a middle man?
I don't. Cursor being a man in the middle between coders and other people models for so long, has so much more training data than anyone else in the world.
There's a plausible synergy in it for xAI though. Access to reams of training data for a company whose marginal cost of compute is very very low, and that they can use as a channel to push Grok. I don't think it's worth anywhere near this much to anyone else, but to xAI it's at least possible.
What matters is that this has enough "future story value" to keep the few investors invested... allowing for the planned index funds to buy into the overvalued stock & allowing for the largest heist in the history of money.
It's become pure hype and drama on the global stock market stage.
Money is neither created nor destroyed in the markets. It's exchanged.
If Musk is worth a trillion while not generating a trillion of value, then it's a heist.
heist: "an elaborate, meticulously planned theft, typically targeting highly valuable items, large sums of money, or financial institutions like banks or museums"
Moat is whatever thing is stopping the next guy from simply drinking your milkshake. People conflate "I could smash this out in a weekend" with "and therefore could also build a multi-billion dollar revenue stream in [counted in months]".
I do think this has had its day. From what I remember, Cursor was useful back in the day when you coded in an IDE and wanted to read code while you baby stepped through incremental changes with an AI. I'm tempted to put /sarc around this but not really...
Sounds like what people here said 20 years ago about Google buying YouTube, or 10 years ago about Facebook buying Instagram - companies with no moats and huge infrastructure costs.
To paraphrase, the biggest trick the devil pulled is convincing founders they need a moat.
In terms of whether they’re overvalued: probably. But any valuation should also take into account the value they have to x.ai (also under SpaceX) as a source of training data for coding models.
Cursor Remote Agents are important to our AI orchestration layer. It's possible that Claude can do this directly but Cursor Remote Agents made this laughably easy.
I guess they're getting bought because they had access to a lot of codebases from a lot of companies, and perhaps there's something to mine in those logs...
AI is good at code but it can't one-shot whole platforms. I don’t think we'll be in the "give me a new Windows XP but Unix core" range anytime soon. You still need time and guidance/ideas.
My honest suspicion is that Musk will focus more and more on AI (and less on space) because he sees it as a path towards his immortality. I expect AI models trained on him combined with millions or billions spent lobbying to allow an AI to own and direct a company. I know this sounds like poorly written sci-fi but I will be only disappointed - not surprised - if post death AI Elon is the richest entity on earth.
AI + Optimus + desire to explore the universe is such a cool concept. I think he'd love to send his consciousness out after all this bickering on earth.
Honestly, the more that we get humans away from squabbling over the same patch of desert the better off we'd probably be. Everyone seems to think they can spend money better than everyone else.
Absolutely unhinged take. Most people just want healthcare, affordable groceries, and a vacation dude. I think Musk has a better chance of being guillotined in the next 20 years than he does sending his consciousness into space.
That'll be true of anyone no matter how much money they have, there's a reason those lines are being sold for votes. There's no end to them and people will always want more, it's not a true goal to work for. Somehow enough people don't understand basic economics enough that they really think taking Elon's "money" (estimated net worth with a million asterisks) will change any of the above for everyone.
They sold at the top. My entire TEAM was weaned off cursor in the last year. New setup - 50% of them do (Codex Desktop, Claude Desktop) + Zed for IDE and the other 50% use (claude code + codex cli) on cmux -- a ghostty based terminal that adds some bells and whistles, literally for notifications when claude is done.
IMHO, the codex desktop app is very powerful for development + testing given it can easily control the computer.
I suspect these won't be as sticky as the contracts you're thinking of. Oracle and Microsoft can run on that because their products aren't compatible with anything else so migrating is a huge pain.
Cursor doesn't really have that. We've got it at $DAYJOB but it's not even the only one, I can also use Zed or Codex or Claude or probably a half dozen other things I don't even know about.
I suspect a lot of companies have that right now because the market for AI is so volatile, and in the near future will trim that down to a couple of tools and Cursor doesn't have much to keep them at the top of the list imo.
I also wouldn't pay 60 billion for a bunch of enterprise contracts that have to compete against Microsoft. No one is dropping Microsoft as a vendor, and they have the ability to up the prices for stuff people need like Office and use that to make Copilot free. If times get tough and companies need to cut costs, it's a lot easier to part ways with Cursor than Microsoft.
Yes. I like it for small fixes, specifying exact code ranges i want touched, and for asking questions because it links me to the code so I can read it.
I use Claude more for greenfield feature building where I dont need to surgically dig in and view existing code specifics.
Yup still got it due to paying for a year up front, its still very decent, especially the newer composer 2.5 model, it's cheap and has a ton of usage included so works well as a general day to day tool.
It’s the best exit for Cursor. There’s not much of a path for Cursor besides getting acquired. Cursor is a fork of VSCode. How much improvement can you really make? Cursor’s own model is based on a Chinese model. OpenAI and Claude are SOTA for coding. The only selling point for Cursor’s model is that it’s cheap.
Absolutely. Imagine forking VSCode and then selling it for $60 BILLION in a few years. And right as you lose your competitive advantage to foundation providers. Amazing.
I don't understand. What SpaceX gets if they stop developing the product - as your comment sounds like it is a dead end? There would be then zero sense to buy it?
cursor feels so 2025 to me guys. these days zed is just way better for my macbook battery and with acp can talk literally to my installed claude code and codex CLI tools, plus their own and custom providers ontop. I was kind of a decade of a vscode user and always just stayed through the evolutions until cursor, but at some point I just need a lean fast editor+lsp combo, git included and a chat pane next to it that uses my real subs underneath easily. (also: codex-cli can spawn and manage subagents and _resume_ them, acting like a real manager).
could be only me though, but longer interactions over days makes my codex gui app grind to a crawl and cursor was not only expensive with opus via api costs but also heating my room a lot. now I have a dozen zed instances open all crunching along with LLMs barely noticeable on system load (except the occasianal testuite runs but thats expected).
If you've ever used Cursor, login, make sure privacy setting is on and contact them to say you want all your data deleted if there is some stored. Then delete your account.
Everyone I work with who used Cursor stopped using Cursor when Claude Code came along. They're back to their regular IDE when the need to read code, or they just review it at PR time. I never used Cursor, but Zed is my favorite editor with an agent. It can use Claude Code, among other CLIs, via ACP, so you can use rolling subscription tokens, or it can use OpenRouter or others if you want a broad spectrum of models. And it's crazy fast. It used to be that Copilot Pro was the best deal on agentic coding with several models from several vendors available, but they've really nerfed it, with uselessly restrictive token budgets and only older models are now available from the major labs. These days, might as well just have a Claude or Codex subscription and use the CLI with ACP in whatever editor you prefer.
I wonder where they will take this, if they'll use the Cursor team to help make Grok Build (which is not just a tool like Claude Code, but an actual Grok model too) more refined for programming? Would make sense to me, and in turn also provide Cursor with more compute they can use.
It also provides xAI with a pre-existing enterprise distribution channel. At the end of the day, distribution is equally as important as the underlying product itself and in some cases is even more critical.
I think that's another thing, while I would like to test Grok more, the Grok AI plans are very generic and not tailored specifically for programming, which is frustrating because I get maybe 8 hours out of Grok for $40 for an entire month, I do wonder if they offered a "Grok Build only plan" if it would actually give me access to more compute. Maybe they intend on making it through Cursor.
I do hope that Cursor doesn't remove any of its current model offerings, and just offers Grok Build in addition to what they already offer, in my opinion unless most of their clients "switch" to Grok (like metrics show they're mostly using Grok vs other models), it would make more sense.
Enterprise AI adoption has reached a point now where FinOps matter, and a harness platform story with a discounted underlying model can be enticing for a number of organizations.
I've seen Gemini land well in a F100 well known for their AI hardware story for that reason, and Alibaba's leadership canned the OSS minded Qwen team in order to build a similar commercial minded approach as well.
At least in cybersecurity, we're also reaching a point where the harness is starting to matter more than the underlying foundation models, and building a harness/bedrock style story while discounting a specific model can play well in upper market deals.
What makes you think so? Is it only because you dont like Musk or do you have some insight into all companies using Cursor you want to share with us? Even if you dont like Musk you should realise that others may not share your sentiments and/or may have similar feelings regarding SamA, DarioA or any of the other CEOs in this field.
Enterprise customers care about things like guardrails and data safety. xAI has always been anti guardrails, and who knows if you can trust them with your data.
At least in the F1000 RFPs I've seen and the decisionmakers I've chatted with, when they talk about AI guardrails what they mean is generic API (eg. can we rate limit, block connections, RBAC/ABAC capabilities, etc) and Data Security (eg. ZDR, encryption at rest/transit, controlled access) controls.
There is a recognition that foundation models and tools leveraging them will introduce some degree of nondeterminism, so the best way to solve that is to leverage preexisting best practice that is used to reduce lateral movement risk by humans (who are similarly nondeterministic in nature).
My company’s security team is very much “no proprietary data or information can be used to train a model”, I just don’t know how you can validate or trust that they aren’t doing just that.
ZDRs, spot audits, and the fact that salespeople can be held personally liable both financially and even criminally for fraud if they sold a contract with a ZDR that was not honored.
They'll likely just take all the various enterprise contracts to help inflate SpaceX earnings, that's the only reason why you'd buy this if you were Musk. Need to help keep the lie going as long as you're still alive.
Cursor's value add as a developer seems much slimmer than the 60bn price tag justifies, but I guess they have a lot of data from the non private usage which bumps the value up?
The product itself is practically a vscode wrapper with Agent implementation and K2.5 forked model (composer).
I’d rather buy 60 “overpriced” Instagrams in 2012, or 3 overpriced WhatsApps in 2014 or 1.5 overpriced Twitters in 2022. I can’t tell what’s a bubble and how money works anymore.
SpaceX is 90% xAI and 10% space. It's no longer a space exploration company, it's an AI company (with a sub-par product). Insane (and dumb) valuation, kind of like Tesla.
For those who aren't aware, Cursor sports one of the best LLM harnesses for coding. The app itself is annoying to use compared to their CLI counterparts, but the harness is widely recognized as the best in the business, or very close. Buying that harness makes a lot of sense considering the cash Musk invested in Grok. He's clearly trying to play with the big boys and grab a chunk of the LLM-assisted dev market.
How does this get a Starship to land on Mars or coast-to-coast full self driving? $60,000,000,000 towards one of those goals would have checked-off one of those boxes.
I'm still waiting for the real news to drop- in the next 6 months we're going to start hearing some big moves from Space X AI. Early this year they lost pretty much all their leadership, it's very clear they failed to keep up with the frontier models and Musk has essentially given up for now - renting out their compute to Anthropic and Google. But that's not sustainable, everything they say about their IPO is that AI is the core value driver. So at some point Musk is going to have some decisions to make about who he brings in to drive that. I imagine once they get that person in and start building a team around them the deals with Anthropic and Google will be ended.
I guess the cursor guys will be happy because they got their pay day, but I'd be very aware if I were them that their future is at the whim of whoever Musk appoints and it's difficult to tell who that would be right now.
I guess now is the time to take bets, so I'm going to bet an early OpenAI employee like Sutskever gets the job and they acquihire him in. Here's a bit of a laugh - at this stock price Musk could probably tempt Demis to come over, that would be wild.
It's a good business. Maybe even a great business. But it's not going to justify a valuation like Space X. In the same way that Tesla has slowly become less competitive in automotive, I don't think it's sensible to assume Space X will have some durable edge in building data centres, especially when basically everything going into those data centres are commodities. And if it is just a neo cloud, then you have to contend with the pro-cyclical nature of that. It's also just clearly not their plan, they're not promising to be a neo cloud. They're promising to own the full stack.
Props to Elon: he IPOs a space based telecom company wrapped in a bunch of garbage companies (xAI et al) as a meme stock and then uses those Elon bucks to go and buy a company to make his struggling ai ambitions relevant.
The formula is: create paper wealth, use that paper wealth to buy something else, and rinse and repeat.
The rich just live in a different part of the multiverse than the rest of us.
>Their expertise will be used to improve Grok Build coding agent.
Is Grok not a toxic enough brand (by association with Musk) that people who would use Cursor wouldn't avoid Grok?
Like, the assumption seems to be that all the goodwill that Cursor users have towards Cursor will now apply automatically to Grok, which seems like a pretty significant leap.
I've been teaching myself physics lately and have found Grok to be one of the best both at coming to a correct answer and helping me to understand how to get it myself. It also seems a lot better than other models at saying "I don't know" or pointing out when my question doesn't make sense.
I'm just trying to get it to help me learn physics. If my topic of interest shifts to mid-20th century European history I'll keep what you said in mind.
If you're going to use the model to learn history you're going to learn the version of history that the model teaches you. A little bit of digging around grokpedia should give you some idea of what that model thinks
I'm a heavy Cursor user, I spend hundreds of dollars per month in overages above my $200 subscription, and I'll be switching away. I have zero interest in Musk's AI.
The CEOs ideology matters due to the fact it impacts the product design. The reason people don't want to use Grok is because it's bad, and it's bad because the team behind grok have to spend cycles crowbarring in far right white genocide conspiracy theories so that it doesn't embarrass their boss on twitter. One of the things we learned with Anthropic is that you have a lot more success being focused on the core product - coding agent, than trying to do that and chase internet chatbot users.
As someone that trains LLMs, even if grok does have a "avoid ""woke""" fine-tuning, adding that needs a few thousand examples SFT and a system prompt. It costs nothing extra to do it to coding agents and is not the reason why grok sucks at coding. It's just not in the same league in general - it's 0.5T parameters only and not trained specifically for coding at all.
Even if the way they are doing it did damage coding performance, it is a simple matter of serving another model without that fine tuning in the enterprise API preferably only to the grok coding harness (or cursor, now). Coding performance for subscription plans don't move the needle in terms of revenue anyways and quality there doesn't matter as much.
Achieving the improvements from K2.5 -> composer 2.5 with just post-training is actually more impressive. Though I believe their next model is trained from scratch.
> So SpaceX is buying 15 years of future cashflows, assuming no change in revenue. Which is bad assumption given past growth was gigantic.
I'd argue it's a bad assumption in the opposite direction. There's no moat. People can and will switch tooling and Cursor could easily be left with a steep decline in users.
But revenue is not really the informative quantity. If you sell gold you will have a huge revenue, but very little profit. I can be a trillion dollar company too if we exchange dollar bills for face value.
"training good coding models" many would say that is a highly debatable statement, and some would say that is just flat out not true. Cursor has not trained a frontier model from scratch, what they did was take an already made (non-frontier) model and further trained it on their user data about coding outcomes from its coding agent. So, a form of distillation and RL.
Sure - why use cash when you can use bits of paper instead?
I'd expect more of the same to come - good way to lock in some of this crazy SpaceX valuation by converting it into something with a bit more inherent worth.
Cursor is massively overvalued. But so is SpaceX so it all evens out in the end.
> each share of Cursor’s common stock and each share of Cursor’s preferred stock outstanding immediately prior to the Effective Time of the Merger will be automatically converted into the right to receive shares of the Company’s Class A common stock based on an implied equity value of Cursor of $60.0 billion and the price of the Company’s Class A common stock equal to the volume-weighted average closing price thereof over the seven consecutive trading days
Current market cap is 2.66T which is pretty bonkers. Thats about intel, amd, and micron put together.
I cant believe people still use Devin. This is just filling the niche of vibe PR'ing? Not even vibe coding? Another level removed from the codebase than even claude code/codex are?
I love cursor. It's so frustrating that large companies are allowed to buy everything. Think of a world where github was still independent, cursor remained independent, heroku wasn't part of Salesforce. All great products that get eroded by the neglect of big tech.
The good thing is that cursor has no network effects. you could make your own (barely usable) cursor on top of vs in a weekend with the features you want.
tbf, irobot being independent killed them. Sometimes it works well for these companies that would otherwise die. I strongly doubt cursor would survive against anthropic in 5-10 years.
I don’t think iRobot died because they were independent. I think it’s more a symptom of them stopping R&D investment because they were pushed to distribute cash.
I was (until this announcement/today) an annual subscriber just for the early days of Autocomplete goodness that I do not use anymore but would stick around and check from time to time their progress and always hoped we would eventually have a good UI for LLMs outside the terminal.
an IDE to look at one file at a time is not interface for LLMs - it was made for people, and while I guess you can sit in the Agents mode all day, to me thats a completely different byproduct and just "Terminal is yuck" kinda users, maybe someday we'll have a proper dedicated LLM UI but we are not there yet.
"In its IPO filing, the company had said Cursor's access to developers' data, including coding requests and design decisions, could help improve its AI models such as Grok."
What are the best open source IDE alternatives to Cursor? There was Continue for a bit but Cursor bought it. Is Visual Code by itself the only open source IDE atm? My main gripe with Visual Code is it doesn't make it easy to use open weight models or non Copilot model APIs. Continue helped but its now part of Cursor.
As for coding in the terminal, we have Opencode, Claude Code, and Codex etc. They are all open source but only Opencode can route to open weight models. But non of these are really an IDE like Visual Code or Cursor.
Are we missing a really good open source IDE to use open weight models? It seems like we are.
I think it's difficult to compete in this space because right now to build the "full IDE" you have to build an extremely capable harness (very hard) and a very good IDE (very hard). I plan to continue using Claude Code and built myself a small tool to verify what the agents actually do + move around the codebase quickly but it's a far cry from a full IDE: https://cotect.dev
I don't know what's more bizarre, either it's SpaceX buying what is essentially a clone of VSCode with AI tools on top, or a VSCode fork being valued billions of dollars.
Not totally, because the deal is in stock. The cash SpaceX got is actual cash that can be deployed today. The stock will be in a lockout period and could be worth nothing or something whenever the lockout ends.
I use cursor (through a work subscription), only the cli (https://cursor.com/cli), and mostly using Composer 2.5, but I freely change the model when the need arises.
Most comments here seem to think there is no command line client? I have never used the editor.
For my personal projects, I use a heavily modified pi. I also have access to a claude code account through work (bedrock), but I don't use it much. It always seems to be down.
I find when working with just a single model (Codex / Claude Code), it tends to have more blind spots. So it's nicer to have different models review each others.
Also with Cursor you can make plans with more expensive models and execute using cheaper ones like Composer 2.5.
This is the best outcome for Cursuor. However, not sure if it's a good investment on SpaceX side though. I remember using Cursuor in early 2025, but after switching to claude code, I hardly used cursuor or even their agent cli tool. The competition is getting quite tough for all the AI tooling and I'm very curious to see how SpaceX is going to do with Cursuor
SpaceX valuation isn't based on ARR so juicing it isn't really relevant. And if he wanted to buy ARR for financial engineering reasons, he could buy a lot more than $4B with $60B
Local inference seems to be catching up and Pi seems to be leading the pack for open weight harnesses. Bold move, SpaceX. I truly hope it doesn't work out for you.
How are these numbers even working out, I get free markets and all that, but Microsoft paid 2.5B for Minecraft, which was printing money at the time (seems they still lost on that deal).
Now a rocket company is buying an editor company for 60B and everyone seems to think that makes sense.
I’m happy to be old man yelling at clouds here because I can’t for the life of me figure out these valuations and purchases.
There's a lot of investment in AI for its potential.
An AI editor company might never make 60B itself, but it might help another AI company grow faster (relative to its competitors, who might also want to buy the AI editor company).
What else can an AI giant do with all that money?
Build in-house: they do, and there's only so fast they can hire/build.
Save? Yes, still do, but if they save it all, and let competitors buy Cursor, they lose.
Invest in other fields? Sure, but if they lose the AI race, that's all they'll be left with.
Tesla's IPO is a bet that if Musk has the right opportunity, he will do well. So he's given a big bucket of money, and needs a team that can deliver. So he buys Cursor.
The winners are Cursor. The losers are whoever is funding the AI companies that get outcompeted.
(Full disclosure: I don't know anything about Cursor, nor much about Tesla or its IPO.)
But surely 60B could buy you something better if you want to spend money on AI. The number seems completely arbitrary, would Cursor say no to 40B?
I really don’t see Cursor as bringing them anything of actual value, it’s more of a bragging thing, but I can be very wrong.
Couple things. First, it was a all-stock deal, not cash, so it's closer to a merger than anything. Second, this deal was determined months ago, most likely at a equity stake defined then. That could mean that a 30B deal 3 months ago is now worth 60B.
Curser is the only possible play to make Grok AI some how useful to enterprise, where the vast majority of spend is / will be? Feels like it could only possibly worth this much to Musk. For me Anthropic are too far ahead - this will likely all be for zero shareholder value when Anthropic continues to pull ahead. Will be interesting to see if Musk switches off Collosus to impede Anthropic at some point.
I am still trying to find a replacement for cursor. But in the past 30D my automations and my own coding has consumed well above 10 billion tokens for less than 300$ with auto and composer 2.5; while building a fairly stable product with 20-30 daily active users. It feels like it’s too good to be true, because I’ve tried with Claude and codex and it just feels so much more expensive.
I'm excited to test out Zed's offering. I'm also going to look at some of the open source agent harnesses, even though I'll have to give up a fixed cost subscription. I suppose now is probably also a great time to kick the tires on self hosted models.
Everybody remember when Zuckerberg told in an Interview in 2024 that human data does not matter that much or more specifically "individual creators or publishers tend to overestimate the value of their specific content". Something along the line RL-Loops are more important.
Hard to square this with that acquisition which seems to be focused on Cursors vast amount of User Data.
Zuck is correct because the barrier to a large user base is not skill but market share which requires substantial effort, luck, timing, and patience.
Nobody can put a dent into Coca Cola because of their market. Better products exist but there is really no way to compete against $5 billion in marketing allowing them to maintain $50 billion.
Cursor is ~1MM users a day. $60k/per user is high but considering this is a stock buy, Space X "made" $300BN in the first day that is ~20% or one day of positive movement.
For Musk (with his baggage) to create or steer that user base would require a significant investment and time. Why not just slap some coupons from an initial bump to acquire the user base, user experience, and IP?
Cursor was simply able to get early access to openai models and get an early lead doing things that are now obvious and done better by many others. Does anyone really want to use a crippled "enhanced" vscode to interact with a crippled version of codex or claude code?
It's a great thing for consumers and businesses to have another competitive, American coding harness + frontier coding model duo. No one wants a crippled version of codex or claude code and surely SpaceX isn't paying $60 billion for that outcome.
I tried it and it was extremely broken and dramatically overhyped. Which I guess is what Elon finds appealing about it, as that seems to be his preferred type of business.
This should be an indication of how valuably xAI sees the training data that Cursor has accumulated, especially with its work on Composer 2.5. Last month, SpaceX and Cursor announced that they had been working on training a new model from scratch. Interested to see if this will put Cursor back into the zeitgeist.
What does cursor have? An ide and coding orchestration? They are using Claude or codex for llms, so they get acquired for their user base and tooling? Feels like a lot of money for that given Claude has the majority mindshare.
I keep hearing this but they’ve been bleeding users since Claude Code came out so a bulk of their data is pre-Sonnet 4 and I’m not sure how the data from users prompting weaker models will help them now?
They did use that data to make Composer 2.5 which was decent but still a step back from GPT 5.5 and Opus 4.8. Though it’s really good at UI.
Not sure how much data they retained but it's possible they also have the _input_ tokens, and in the "early" days (how far everything moves in just a couple years...) more of those input tokens would have been non-AI-generated
They technically have their own models as well, but they are based on other peoples models.
Most of this is just customers + staff/tech rather than models being acquired. Cursors actually got so much better in the last year. Their composer 2 model (a tweaked version of Kimi K2.5) is decent for day to day mundane tasks and the app can auto switch to more capable models when needed.
They have millions of conversations with code prompts and diffs and the user telling them the model is taking a wrong turn. This is quite valuable data that is probably already used to tweak Kimi code (composer 2).
Moreover they have quite the enterprise client base that I suspect at least a portion of which will jump ship with the acquisition.
People keep saying this but I have yet to find an integrated system that has good tab completion, cheap coding models and works well in an IDE. There are a number of options out there, none of them have captured much market share.
Have tried it over time. I think it’s decent but I still found it to be not as great or quick as Cursor and the rest of their AI integration did not feel as compelling for my workflow.l but overall agree.
Do you have a point? The whole package is compelling like I already said. Nobody has replicated it but everyone likes to ask what is the value but never share a compelling competing product. I am not Boris so I still write some code or make manual tweaks, autocomplete is nice.
Edit: I realize my question is maybe harsh but I think it’s valid for these drive by comments that drop a question like “isn’t tab completion dead?” Without any other substance it is a huge detractor to the comment quality of the site. At least add more substance or opinion.
All good, no offense from my side, apologies for the disturbance if it caused any. Generally would have just been curious if people still use tab completion much? Virtually all people I know have switched to Claude Code or Codex. I also read recently that Boris Cherny (CC creator) deleted his IDE. Of course that's marketing, but it shows the current state quite directly. So I'm wondering what your dev workflow looks like? Any why/when you use tab completion instead of coding agents.
No issues here but I think it helps to add a few thoughts to continue the dialogue.
I think it’s largely shaped by the type of work you are doing and I am not sold on the idea that we are at a point of never writing code. I leverage these tools a lot and there is still nuanced mistakes LLMs make that I have to intercept. I am still in the code at least for now. Other than my friends that work in startups I have not seen too much others delete their IDE and not have any touch on the code.
The have customers actually willing to use and pay for their service.
In a world where even Microsoft is needing to use AWS for capacity beyond Azure, xAI's utilization of their data centers has been so low that they are renting them out to competitors instead.
Nobody wants Grok. If you aren't using GPT-5 or Claude, you are probably using an open Chinese model like Qwen hosted by some provider.
I would expect Cursor to be forced to use the Grok Code models in short order. We'll see how people feel about "Mecha Hitler" writing their code.
Customers, talent, training data, an increasingly competitive coding model and now a fuckton of compute.
So, you know. Couple of things.
Specially given that coding turns out to not be all that complicated, in the grand scheme of AI things: I don't think it's going take much more advances at the frontier before code writing will be as good as it need to be. At that point Composer (their model) catches up, what, 6 month later and they good.
Musk bought Cursor for its users (lots of devs still use it), as part of yet another attempt at catching up after building Grok and buying OpenAI failed.
Likely, Cursor becomes Grok Desktop or whatever, and eventually uses xAI's coding model if they can make a competitive one.
I wonder why IDEA didn't catch up with its own agent support and even their own models. It's not like agent harness is that hard to build, right? Or maybe they did, it's just that they haven't won the hearts and minds of the developers like Cursor did?
I think we should get used to it because that's what's going to keep happening again and again. First Twitter, then Cursor. When someone falls behind in the race for innovation, they usually buy the best product available and use it to get ahead of the competition.
Back in my day, a software company of national importance would be considered a big success if acquired for 60 million... Of course, at that point it would have to be profitable, and actually own its key IP...
That's interesting I was using Windsurf before and I really enjoyed it. Then it became part of Devon. I was less thrilled about that so I was looking at Cursor, but now it's also getting bought out. Any suggestions on what else is left? : )
My honest read is that, having everything — the data centers, the compute, the models (however misaligned they might be), the only thing xAI is missing is users. They don’t have any users because the only people who use Grok are essentially Elon’s fanboy club, and all they pretty much do with it is ask it to generate arguments to win their Twitter threads or nonconsensually unclothe people. Cursor gives xAI a captive audience of users; most sophisticated users don’t use it anymore, so anyone left is unlikely to be opinionated when models are shifted to Grok. Marriage made in heaven.
>the only people who use Grok are essentially Elon’s fanboy club
that's not true, Grok compelling feature is it's capabilities, performance and price. You only get comparable prices with GPT-5.4 mini or Gemini Flash.
Also Grok Voice and images are pretty good.
I think at this point they don't want to be at the same level as Opus or GPT, they found their niche.
You generally see this sort of thing in the games industry when a publisher gives a developer a bunch of money to make a game but then the developer screws up and runs out of money. Publisher buys them to try to recoup their investment.
I run Claude Code in cmux with Soonpatch and that's it. I tried looking at Cursor but honestly it wasn't providing any value and I prefer being 100% up to date with CC updates/interface
I’m happy for the Cursor team, but I sincerely don’t understand the valuation.
Most people I know who figured out how to use Claude Code or Codex directly get much more compute and a much better power-user experience. The difference is not even funny. And it’s not an IDE vs terminal because both claude and codex work just fine in vscode.
What annoyed me most was my own experience: last November I paid $20 for Cursor, burned through the quota in one day after just a few prompts, and didn’t find any clear way to even see quota, so i contacted support. Somehow they claimed I had used around $70 of compute and implied I should be grateful. But they calc price of all input tokens as if they were cashmiss tokens, which is obviously more expensive than real API usage and this was extremely dishonest on their part.
So from my perspective, Cursor or similar solutions often look like a middleman between the user and model provider. In theory, that should be hard to defend against using the frontier labs directly.
That is why it is mystery for me why so many people still pay for it over orders of magnitude cheaper claude/codex
Good for them. I stopped using it abruptly once I joined the claude cli + vim train but I am sure they will be happy to cash out. I think a lot of other devs stopped using it too so good timing.
I fail to understand, cursuor is bad at C++, really Bad. In fact of all tools and LLMs, cursor is worst, the best is kimi as per me. It know how to spot segfaults in long lines of code. Why would anyone pay for this hallucinating garbage, especially from a company which does mission critical C++
And, No It;s not a skill Issue. And No, my goal is not to make another vibe code music wrapper around ffmpeg, and No My goal is not make a program around some tcp wrapper. My goal is to fix and write new features in decade old code bases running across millions of devices.
Ugh, I'm already tired of seeing ads everywhere for Cursor about how you can build EVERYTHING and solve all problems using their agentic IDE, so now I have even more reasons to dislike SpaceX too.
I realized a while back that Elon Musk isn't Iron Man. His superhero (or supervillain depending on your view) persona is ZIRP Man, the master of riding successive credit expansion and speculative waves. It's sort of ironic that he at least pays lip service to some Austrian-style quasi-libertarian economic ideas, because the Federal Reserve created him.
Now he's surfing the AI wave. We are no longer technically in ZIRP but the delayed inflationary wave is now traveling through the economy and pumping everything. He knows the best way to soak up cheap money right now is slap AI on it.
I also had the thought the other day that him hitting $1T technical net worth might actually be a harbinger of a lot more future inflation. Inflation of this type hits assets before it hits things like prices and wages, and it hits assets with fast market cycles like stocks before it hits things like Real Estate. The blast wave starts at the top and moves down and out. So maybe Elon hitting $1T really means that in 20 years that'll be more like $100B inflation adjusted. Meanwhile a loaf of bread will be $20 and a starter home $4M.
But the fact that monetary inflation starts top-down is why low interest rates exacerbate inequality. The very richest and most leveraged can use the arbitrage gap to buy everything else before the inflation wave propagates. We've been in a low interest rate environment for about two decades, and you can see during that time how the super-rich with access to cheap money have fully detached from the rest of the economy.
In other words: the reaction to the 2008 financial crisis was to inject huge liquidity at the top, which created the new Gilded Age.
Ultimately it may be somewhat intentional. One way out of a sovereign debt crisis when you also have a sovereign currency is to inflate your way out, which basically is a huge tax on every non-domestic entity that owns your debt.
To be honest and objective, I think he at least knows enough about engineering to hire people who know what they're doing, which is how he got here. It's not all chicanery.
Over time, though, I think he's drifted away from his original "make real things in the real world" focus and more toward "play money games" and "play political games."
It's sad. One common comic book supervillain arc is to start as a hero and become what you despise.
I really hope Musk doesn't ruin the product itself. I am not a fan of how he changed X at all. Id really like to see them stay on the current path, which has been brilliant so far
I had high hopes in Antigravity and Gemini, but they royally screwed up everything to such a level that the only worth plan is to use free plan for creating docs.
Will the $60B be newly minted SPCX shares? Does this effectively dilute investors by about 2-3% (minus some fraction representing the fair value of Cursor)?
Spcx is so overvalued that its value for the investors is whatever they wish it to be. If they wish it causes no dilution their wish becomes true. They already wished about 90% of the whole share price into existence.
Good hopefully Grok Build gets better and they start to innovate in this area.
Claude Code and ChatGPT need more competition. So that they also innovate more.
What are the best open source IDE alternatives to Cursor? There was Continue for a bit but Cursor bought it. Is Visual Code by itself the only open source IDE atm? My main gripe with Visual Code is it doesn't make it easy to use open weight models or non Copilot model APIs. Continue helped but its now part of Cursor.
As for coding in the terminal, we have Opencode, Claude Code, and Codex etc. They are all open source but only Opencode can route to open weight models. But non of these are really an IDE like Visual Code or Cursor.
Are we missing a really good open source IDE to use open weight models? It seems like we are.
Why don't they just make their own in-house development environment? Cursor's codebase is maybe 5,000-50,000$ worth of actual code, even just $10M (compared to the $60B) could knock Cursor out of the park with a completely custom code editor, and even with a smaller budget they could fork VSCode. Maybe for the social value? I feel like a $10B advertising budget for a bespoke AI development environment is more than enough to become an actual competitor.
Yes, it’s my daily driver for building the saas I run full-time. I’m not happy about this news.
I like the ability to switch between any models, Composer 2.5 is really solid, I like having my agents coworking in the IDE with me, the plan mode is great, Cloud Agents are great, especially with slack, linear, web, etc integrations. I routinely tag an error report in slack and Cursor fires up a Composer 2.5 cloud agent that has readonly db access, access to error reporting, etc, and it can triage the issue, issue a PR, and tag me in slack.
The only thing I’ve felt like I’m missing out on is the subsidies of the CC/Codex subscriptions, but it seems like that is rapidly eroding anyway.
It uses AI to replace a very niche human-powered workflow in a niche industry. That's all I'll say, but it's grown to about $30k MRR in the last year and is supporting my family, so the stakes feel pretty high to me.
I actually acquired this as a little side project from some guy online. I hadn’t even really heard of the niche, but I knew enough to know that it had some potential, and I could also see that it could be used for an adjacent market segment that I was pretty familiar with, because my dad worked in that segment when I was growing up. And the price was cheap because it just had a few users and was pretty new. All but one of those users cancelled in the couple months after I bought it, because it turned out the product just didn’t do what it claimed to. But I didn’t really care about that, because 1) I knew it could deliver if built correctly, and 2) I could just tell there was demand for this and it wouldn’t be that hard to sell.
So I turned off the signup form and started rebuilding it and improving it, but I kept getting people emailing me and wanting to sign up. No marketing, nothing. I held them off for almost 18 months while I rebuilt it on the side of my day job, and that gave me a ton of confidence that I was on the right track.
Then about a year ago I got laid off. I was really close to relaunching it at that point, so I used all my severance to go all in on it.
Not to my knowledge? But even if that's the case, Cursor seemed like they were doing fine without SpaceX, and I'd like to avoid giving a single cent to Elon Musk. You can do as you wish.
CC is mostly my default for large tasks / features (ex. Plan > execute plan ) Biggest gripe with Claude Code is that it is painfully slow relative to the other two.
Cursor for small stuff like bug fixes since it has a lot of models to choose from. I love the review/ diff / checkpoint features. It's planning feature is on par with CC. I'd probably use Cursor as primary driver if it had better cost efficiency. Next version or two of Composer may fill that gap in cost/quality/speed.
Codex isn't allowed at my work, but I use it for personal projects. It has the best balance of quality / cost / speed even if it's planner is poor and quite frankly the codex harness needs to catch up with the other two.
CC for quality / cost. Cursor for quality / speed. Codex for balance of the 3.
composer is competitive with around opus 4.5 in feeling?
largely lags behind opus4.7/gpt5.4, but is respectable, and generally outperforms the glm/qwen equivalents anecdotally despite benchmarks.
fails to follow instructions more often, and is less code critical, but performs okay if you can decompose the task to smaller problem spaces. i.e. only do manual review, only do typechecking, only do specific component. etc
I agree, Composer 2.5 is really good. I use it for all kinds of small tasks, and really for any kind of first pass at debugging, answering questions about the codebase, pulling data for reports, etc. It’s fast, pretty accurate, and basically free.
Yes, I still use it, although less than I would otherwise.
Good:
- Composer 2.5 is pretty decent for the quality / price ratio.
- Easy to assign an issue to it in Linear (I know Linear just added this natively for linear agent, but it seems rubbish compared to Cursor)
- Bugbot actually finds some useful issues (things Claude and Codex will miss)
- Using @cursor in github usually works well, and better than @copilot.
- Working with Python Monorepos with UV in their IDE. VSCode and Cursor work well here (Antigravity managed to screw it up somehow).
The Bad:
- Usage/billing dashboards - These are are opaque and you can't attribute what actions map to what spend.
- cursor won't follow PRs well like Claude Codes does.
- Setting up environments is less good than Claude Code
- Their IDE fork is woefully out of date, it'd be nice if it had more of the codeium fixes.
The Ugly:
- Settings - Try to turn off bugbot, there's multiple places you have to do it. Good luck figuring them all out.
- Support - they are polite, but gas light you and tell you it's your fault their product's settings are awful.
Linear employee here - if you have any specific feedback on our Claude/Codex integration, happy to hear it. Definitely a v1 so expect a number of fast follows up with some of the missing functionality like env customization, secrets, and code signing.
I was speaking more on the linear agent vs the existing integrations. We love the linear guided reviews and issue tracking so have high hopes on getting a good DX here from y'all.
The claude integration - works as well as anthropic will let it work, since you can either automate it (anywhere from 'claude -p xxxx' or the api and ignore your subscription and pay by the token, or open a crapton of tabs with the terminal, or paste it into a bunch of sessions in their app. Which works more-or-less but it's cheaper than per-token costs.
The linear agent, doesn't seem to read the AGENTS.md file, follow along on a PR nor nor let you configure a sandbox (it told me this:
```
Note: I couldn't run ruff/pytest here (no uv/venv in the sandbox), so I verified syntax via AST parse only. The Postgres-backed tests will run in CI.
```
After I asked it to look at the PR check failures.
To be fair, claude code does it 70% of the time (the other 30% the sandbox is dead), and cursor about 10% of the time.
I’ve heard of space based microwave energy weapons. I haven’t looked at the physics personally but I talked to someone who did.
Something like a 10MW phased array to create a 1 cubic meter ball of plasma in the atmosphere. I wonder about the transmission losses… but what a weapon that would be!
meanwhile all I need is a cursor (lowercase) and I can program and do all the actual software engineering I need. been true for many decades and likely will for many more
cursor acquisition and anthropic fable models banned from export by US government cant be coincidence. to keep powerful ai tools and model brains onside country
Ai is great. The bubble will burst. We will keep Ai just like we kept "the internet" after the dot-com bubble, but we still won't buy our pets online. I mean London has pretty good train connections and stations because a bunch of companies repeatedly tried to get rich. Most eventually failed, but we kept the rails. I just hope we get our computers back after this round of gambling.
I’m worried some of the people leading this race will try to entangle the institutional financial giants or even the government to ensure some sort of too big to fail scenario.
There’s an unprecedented amount of money at stake. And the admin has never been so openly and blatantly for sale.
What are the alternatives to Graphite these days? Github's stacked PR support is still immature, AFAIK. I would love to see Linear move into this space and start offering both git hosting and stacked PR management.
I mean, I guess SpaceX has a lot of money to throw around. I just don't see how Cursor is worth even 10% of this number. There are so many similar tools out there - many plugins to VS Code (Kilo, Cline, Roo, etc), many CLI tools (claude code, opencode, plus everybody and his brother is putting out their own...).
I believe open source models will win here, mainly because China won't allow otherwise. I also think that nobody is really talking about the hardware decpreciations coming in the next few years, which is going to be really important from a performance-per-Watt perspective. B100s aren't going to suck. But a theoretical T100 will get 30-80% more performance for the same energy input.
So, SpaceX. I've previously said that SpaceX would've been a significantly better company without xAI. SpaceX was used to rescue Elon and other "investros" from the financially disastrous Twitter purchase. Starlink, Starship (which is a risky program) and the Falcon 9 are a solid business. They're just not a $2 trillion business.
So I believe that the AI bubble contributes at least half of SpaceX's valuation and when and if that bubble bursts, at least half of SpaceX's value is at risk.
Google announced they're throwing billions to rent GPUs from SpaceX. That might sound good. It solves a short-term cash issue. But as another commenter put it, it makes SpaceX seem more like a Commercial REIT. After all, renting out your GPUs is literally the lowest-value thing you can do with them. You're not building a business. You're taking rent so someone else can build a business.
So buying Cursor and I'm sure any number of other AI startups in the coming year or two, seems aimed at kicking that AI can down the street.
So I view the Google-SpaceX as a red flag in the short-to-medium term. SpaceX simply can't seem to do anything valuable with all the compute they have. And I also have way more confidence in Anthropic (in particular), OpenAI and Gemini than I do in Grok.
I don't understand your point here. OpenAI rents all of its compute. It doesn't own datacenters. Any advances in hardware or optical computing benefit them because they can serve more customers. There is huge demand for GPT 5.5.
I honestly don't get why they feel the need to buy Cursor or why OpenAI wanted Windsurf.
If it's data you're after, wouldn't it be so much more cheaper to just hire a dedicated team to fork VS code and integrate your own model and give it to the public for free with unlimited usage for a couple of months?
Till the minute I clicked on this I couldn't decide if it was satire or not.
But also, for a few weeks periodicly I've been wondering what's going on with Cursor. Haven't thought about them at all, let alone used them, in quite some time.
They were a pretty big AI-native player. Seems clear we're well into the consolidation phase of this economic cycle.
Is this one of the situations where one of the engineers at SpaceX showed elon musk cursor and said "this is the future of programming" and that has been stuck in his head ever since? Otherwise (unless someone can prove me wrong) I don't quite follow buying out a company at the peak of their evaluation.
This truly shows how king making in venture capital is done, kids have the MIT pedigree but sometimes this is not enough for certain demographics, give them a ton of money to explore ideas and pivot, product is a vscode fork that sells subsidized AI, only possible with venture capital. Providing inference at unsustainable rate deemed as "product market fit". Product loses money until they exit.
VCs that say "I always knew the team was special", give me a break.
Glad I've jumped ship to Zed+ACP. I liked the idea of the new agent window view, but it's obviously slop and hasn't had a any polish or care put behind it given the number of bugs I keep seeing.
Zed is so much more stable and sleek and the agent view (threads) actually integrates nicely in a real editor. The side editor in the agent window was so much worse than the vscode base I expected, I have no idea how they dropped the ball so hard here.
At this point, money and rationality does not make sense to me, rather beyond my ability to understand. But I feel it is all about accounting and writing off eventually and a few profiting from it, not the retail investors for sure. Again I am just saying what I think and there could be no rational to my thought process.
SpaceX should rush to acquire as many companies as they can with stock. The market cap is absolute insanity (I know people keep saying this about new high scores in unrelated-to-reality valuations, but this one might just be the pinnacle), with zero rational basis, and they should try to make it real as rapidly as they can.
Do you really think so? Like everyone who risks their and their clients' money here is just being irrational? Is this really a coherent view? Could it not be that someone knows something you don't, or does not have the biases you have?
Everything was in the S1 filing. There is no "secret knowledge".
The rational basis is entirely "I can sell the stock to somebody else for more money". Where in normal stock it would be "this company can make a profit that gives a return on this investment." This is a purely speculative play.
Why is something true for a company where it is true and not true for companies where it is not true?
SpaceX is worth more than Amazon.
Amazon has $750B of revenue and an enormous unfathomable moat across many fields, across most of the planet. It had a profit of $77B last year. And people consider Amazon overvalued, and a symptom of a serious bubble.
SpaceX has $18B revenue, and a consolidated loss of $4.9B. It has basically zero path to real profit, and it turns out that space launch actually isn't a lucrative industry, so much so that SpaceX had to create its own customer.
I mean, the biggest news about SpaceX has been the utter failure that xAI has been, reducing it to renting out all of the GPUs that Musk forced his other road-to-failure company, Tesla, to hand over to the failure that was xAI, that failed so badly it got hidden inside SpaceX. Good god. Somewhere in there the failure that was SolarCity got packaged in the giant scam.
Like seriously, the biggest win of the company is that it absorbed the husk of xAI that had a massive surplus of GPUs from when Elon tried to buy himself credibility in AI, and the market happened to make them worth more so that's their big win. Their biggest success is basically scalping GPUs.
>even though literally everyone hates him?'
Guy basically runs the US government, which has been reduced to a banana republic plutocracy. People invest in him because they know the system is so catastrophically corrupt that he'll always come up a winner, regardless of how enormous of failures he keeps generating.
I mean, this is such a transparent shell game scam that they're immediately talking about packaging up even more of Elon's scam businesses together. Just amazing stuff.
Don't worry, those super robots are coming any day now!
Yes, the US market is in the end games of a massive spiral -- a circle jerk of trillions of fake dollars moving in a rapidly accelerating circle -- and it will not turn out well. SpaceX is the moment when it is laid bare.
> Guy basically runs the US government, which has been reduced to a banana republic plutocracy.
Do you have any evidence of this? As far as I know, he basically never gets what he wants. He was against Trump's Big Beautiful Bill, it passed anyway. He wanted DOGE, it fizzled. He wants more Solar/Electric car related subsidies, Trump does not give a darn. He wants more H1Bs, Trump has been doing everything to frustrate H1Bs.
The one thing which Musk seems to have gotten was Jared Isaacman, but that was really difficult for him to get, and it took way longer than it should have.
Really difficult to see evidence that he runs the US government.
Point us at a rational verbalized or written argument for SpaceX's current valuation (and increasing)? Everything I've read says the valuation is too high and here is why, with x, y, and z reasons. Everyone I read who encourages buying seems to be ignoring arguments entirely and going on vibes.
Interesting. I've not used Cursor in almost a year after using Codex/Claude.
Won't be surprised if Elon paid another hefty premium.
I just realized this whole game is just getting rich from other people's money and there might not be people left to buy those people's shares when the music stops.
I can't stand Cursor.
Every time I open it up I have 3 popups I don't use, that I then need to figure out how to close.
Using it for notes is impossible, since the autocomplete just tries to fill in bullshit.
Awful what VC money did to it. Hope to never use it again, now that work stopped mandating it.
I quite liked Cursor. I even tried Claude Code and found myself wanting to go back to Cursor. Unfortunately, this completely kills it for me. I will not support Elon Musk or any of his shenanigans. He is already far richer than any person should be, but he also constantly tries to manipulate the government to benefit himself to the detriment of everyone else, whether that's DOGE, or the fast track to begging added to indexes on the stock market, or burying all the investigations into Tesla. I cannot in good conscious pay for a product when I know that he is profiting from it. So long Cursor, it was a good ride while it lasted.
Another grab, I wonder there has to be something that someone pays $60B for, but my question is: I often struggle to understand why someone needs it. Is it extra plumbing and easement around code development or paying base?
"In its IPO filing, the company had said Cursor's access to developers' data, including coding requests and design decisions, could help improve its AI models such as Grok."
> With the option agreement, we have the right, but not obligation, to acquire Cursor at a predetermined price or pay a fee
> The consideration for the acquisition of Cursor, if any, after the closing of this offering would consist of shares of our Class A common stock based on an implied equity value of Cursor of $60.0 billion
Pieter is so dumb. All he seems to do is post comparisons between the wonderful U.S. and dying EU that are completely wrong. If Elon is listening to Pieter, pray for Elon.
This is true if you take the ai market as equal to the market for labor discounted to 5-10% penetration.
It’s not a totally unreasonable assertion, it’s the implication of the assertion that we are uncomfortable with. There is no reason for the models to stop their improvements in the near future.
> There is no reason for the models to stop their improvements in the near future.
Sure there is.
1. The cost of each new generation of training runs appears to be rapidly rising
2. The Trump admin just told the leading model to stop making it available to non-Americans, which in practice meant stop providing it at all
3. The factories to make the hardware are hitting bottlenecks, and while they've currently been navigated around, there's never a guarantee the next one will be
Currently I'm wondering at what point the direct impact on the US energy supply gives the US a taste of Baumol's cost disease as AI companies continue to outbid everyone else for electricity.
There are some counters to this, especially in electricity. We'll see massive expansions of wind and solar in the US because of this. Both the speed of install and low costs will guarantee it.
> We'll see massive expansions of wind and solar in the US because of this. Both the speed of install and low costs will guarantee it.
Implausible while Trump remains in office. He hates renewables, shuts them down even when doing so actively costs money.
Between AI hallucinated content and the politicisation of the numbers, I'm not sure how much AI compute capacity is being planned right now; would you accept a claim of 300 GW? It's a number I heard recently.
Given the capacity factor of PV, even China would have to think carefully before supplying that much PV over the next few years (300 GW avg ~= 3TW nameplate).
(Not sure about wind, wind's CF seems to vary between years).
Unless it's something on federal land businesses are pretty much ignoring Trump on renewables.
And 300GW power planned doesn't seem too far out of bounds, there are a huge number of 'planned' data centers all over the US.
World wide over 800GW of solar and wind was installed in 2025 and 2026 numbers should be over 1TW of renewables. How much of that will the US install itself is a much smaller percent, but as power prices increase the pace to profit off of it will quicken. I know China installed over 300GW themselves last year.
> Unless it's something on federal land businesses are pretty much ignoring Trump on renewables.
Except for all of the tariffs etc.
And that's without Trump seeming to be actively choosing winners based on favour to him, as with supporting Grok despite the data centre pollution in another thread and *possibly* (I don't wish to overstate my case) the ban on Fable.
> World wide over 800GW of solar and wind was installed in 2025 and 2026 numbers should be over 1TW of renewables. How much of that will the US install itself is a much smaller percent, but as power prices increase the pace to profit off of it will quicken. I know China installed over 300GW themselves last year.
This is why I wrote:
Given the capacity factor of PV, even China would have to think carefully before supplying that much PV over the next few years (300 GW avg ~= 3TW nameplate).
When the recent good news is that "the world installs 800 GW of PV" (TBH, I thought this was closer to the last 12 months of just PV than the sum PV+wind), that's the nameplate capacity, not the actual year-long-average output, which is about a tenth of that.
The most recent PV capacity factor number on Wikipedia was 13%, which would make "800 GW" only 104 GW in real output; the figures I see for wind are that the CF is 25% (with much higher variability) but the nameplate capacity is lower, so they're pretty close as totals in real total currently installed output.
As long as Chinese companies keep pushing on, so must US companies too.
It would not surprise me at all if we suddenly start seeing top US AI companies lobby against Chinese models, or even the gov. making it illegal to use Chinese AI models.
But in this day and age, I just don't think it is possible. A distant third option would be that the big AI companies try to make hardware so expensive that people simply can't run their own models, while blocking access to foreign models.
> There is no reason for the models to stop their improvements in the near future.
You speak as if "improvements to models" is just function of time, and resources are infinite.
Models keep improving as long as there are resources to allow for larger and larger datacenters, if we hit a scientific breakthrough once LLM technology become the bottleneck, if the economy is infinite to allow infinite growth, and (geo)politics is not a thing to worry about. Or we discover ASI, machine improve themselves and we reach the technological singularity.
I know everybody is drinking the kool aid by the gallon, but can we maintain a little bit of objectivity?
The s curve won’t inflect until it becomes difficult to allocate additional resources due to economic limitations. There is no sign that training a model on 10x the compute won’t lead to at least an equivalent improvement as the last order of magnitude increase.
If we define the Pareto frontier’s input in terms of a magic “compute equivalent unit”. We get a free order of magnitude from nvidia hardware improvements every 2-3 years. We get another order of magnitude from capital expenditure every 6-12 months. Kernel improvements to the models themselves likely yield an order of magnitude gain at some periodicity.
> the market for AI is 2.5 larger than all the food sold on this planet.
It just shows how much the automation has impacted agriculture and the food industry. Sure, there're rural farms that apply 200 yo technology. But e.g. the grain production and farming are incredibly efficient at scale. So, it's not that costly for as a humanity to feed 8 billion people (at a varying level, of course).
Food is a solved problem. We can grow far more food than we need and we stop doing so simply because the low prices mean it's not economically viable.
In the places where famine remains a problem, it's due to political issues, not that we can't grow enough.
And growing all that food requires a tiny workforce compared to 400 years ago before the Agricultural Revolution. AI might extend such a massive reduction in labour requirements to many other industries.
> We can grow far more food than we need and we stop doing so simply because the low prices mean it's not economically viable.
Half. This depends on there being a reliable source of cheap fertiliser, which would be much more secure if not for the situations regarding Hormuz and Russia.
> We can grow far more food than we need and we stop doing so simply because the low prices mean it's not economically viable.
So, it's not a solved problem.
Last time I checked we have plenty of people in several parts of the world with difficulties to access the required level of food to be healthy.
Technologically yes, but this is a vast oversimplification.
You need lots of money to be able to buy the tech you need to do so. And you can't exactly earn that from not using the tech, since foreign (or even local) competition will slaughter you on prices. And if you do make it, you're stuck with a low-margin race to the bottom on price.
I don't think anyone is claiming AI and food have the same elasticity of demand, which is what this really talks to, but, after a claim the AI market is 26 trillion dollars... I wouldn't be surprised if someone did.
Ackshully, AI does have an upper bound in information theory, but since we're not anywhere close to writing data to the surface of a black hole I don't think it's a big issue yet.
We could remove 100% of world AI supply and humanity would not be worse off. It is still additive and in areas of generally indeterminate value except in hype.
Reasoning and RAG is amazing already and is a productivity gain but I'm yet to be convinced GenAI is anything but a slop machine.
You can't food maxx a trillion calories a day to generate a multi million dollar bill. You can token maxx it though.
I think the issue is the reality that most life is worth a lot less (in US Freedom units) than some software running doing absolutely nothing truly valuable for anyone.
What precisely is the moat surrounding AI that SpaceX is using to justify this kind of spending spree? I don't how SpaceX and other AI companies will be able to keep the weights of their AI models private in the face of interest by virtually everyone in the world. It would be absolutely trivial for a nation state to walk into a data center using a state issued security certificate to seize a few of the physical servers running the cloud services of OpenAI / Grok / Claude. Copying the weights is trivial. Infiltrating a company with spies as new hire coders to gain access to source code is also trivial.
This is really starting to feel like the pets.com era again.
This is an insane take. Of course there are other areas that are growing or could grow if there was investment.
The problem is an absolute lack of vision on the part of those holding the capital.
Meeting the challenges of climate change could hold huge opportunities. Look at China’s massive expansion in renewables, look at the expansion of renewables in the US despite political headwinds.
Have some imagination, break out of your echo chamber. AI ain’t the only game to be played.
It's not lack of vision. It's that capital demands gambling. Everyone knows you could plough money into big projects in the US and double your money reliably. But the powers that be do not want that. They want to gamble, and try to become trillionaire #2.
It's not insane. The GP is correctly describing a bubble economy.
The money chasing investments is orders of magnitude larger than the money people have on their pockets to spend. As a consequence, the only profitable thing to do is sell capital goods to make business and there is no profit on selling actually useful things.
China is in a different reality in large part because of their capital barriers that stop money from flowing in. Countries with bad reputation are also less affected.
What the GP gets wrong is that none of this makes AI a good business. Instead, it makes Nvidia a good business, but that's not news.
Looking at venture funding, it's definitely true. That doesn't mean other problems don't exist or aren't worth solving. But the concentration of (competing) capital and talent is insane.
Not to be a jerk or anything, but that only matters for the guys and gals left holding the bag.
I don't think that will be Musk. He'll probably pull out significant resource from all this financial engineering relatively quickly. Probably via more financial engineering.
Cursor is a harness that can be used with all kinds of models. It's a much better harness than anyone else's and takes the company out of just playing the model game.
Exactly this. I think valuations and the AI market could get stirred up if:
- We get an open source Opus 4.8 equivalent and pair it with an open source coding agent
- Running this OS stack becomes cheaper than what frontier model providers charge (see OS model prices on OpenRouter vs. frontier lab prices)
- This happens across verticals (i.e. not just software)
The first “DeepSeek moment” didn’t do much damage back in the days, but I wouldn’t be surprised if a similar moment becomes a lasting, effective, cheaper alternative.
OpenCode exists, it is your "open source coding agent" that is practically on par with Claude Code and Copilot in terms of being able to do the 80% of things that most people actually use.
DeepSeek v4 Flash/Pro also exist, they are open weight and on par with Sonnet, just a bit below Opus. Again: practically useful and sufficient for 80% of things most people actually do. And most of the remaining 20% are benchmarks designed to push the limits, not productive work.
Using these already is way cheaper than your typical Claude API prices. What's still missing is a) mindshare - everyone still thinks "claude = coding" and everyone thinks he/she really needs the very best models because he/she is doing such incredibly complex stuff - and b) someone pushing such a stack as a convenient solution for corporations to easily dump their token money into, complete with user management, enrollment, monitoring, all that enterprisey stuff you need if you want to sell to, well, enterprise customers.
Need to give that a try then. What's your experience so far? Also in terms of cost/benefit?
My current hypothesis is that the $150 Claude Max subscription - of which I barely hit the limits anyways, even though I used it non-stop at work - still is very cost effective.
If the price of Claude Max increased significantly (say 2-3x), and my business would balk at paying the subscription, then I'd look for an alternative.
They are a distant third at best, at least in trading companies. If you look at Chinese and other likely national actors, they are probably further down.
The thing with dotComs is that they didn't have THIS level of unsustainable financing burn, and a tangible issue of token processing cost that has no magic wand coming with the current practical limits of Moore's law.
To be fair, food is the smallest bucket of my monthly expenses. And there are many people here on hacker news who pay more for their AI tokens than for their food.
How does argrar industry and tech industry compare as share of gdp in the US?
We'll figure out how to make it much much cheaper to produce the compute needs we have today with tokens. The question is will how many new use cases arise that need much more - clearly we aren't meeting needs (price stays high) but how much more do we need, 100x, 10x, 10^6 x?
Look at electricity, the world of 1900 could not create enough electricity or even conceive of how to add enough to meet 1950s needs. But we made it incredibly cheaper to produce, but also created a lot more, and boy do we have so much more use of electricity now. And it's not that expensive for a human to pay for their needs (not free, its not cheap for poor people but it's still gotten cheaper).
Food is worth a lot more than that. If the alternative was starvation, we would pay approximately all the money for food. By that metric food is worth more than $100T. The difference between $100T and $10T is called the consumer surplus, one of the largest benefits of a free market economy.
AI might eventually provide $26T worth of value, but if it captures anywhere close to that amount of revenue that'll indicate a failure of the free market economy. Competition and open source will have failed and the oligarchy has won.
(Either that, or inflation will have made $26T a relatively smaller number).
Always fun to remember when calculating TAM - something like 85 - 90% of the world earns less than $1000 usd per month.
The math don’t math here, there literally aren’t enough people to afford this and businesses will go under the more people are displaced for gainful employment.
There have been a few recent stories about businesses finding themselves spending more on tokens than they were spending on the workers these AI Agents were supposed to have replaced.
You need to understand the definition of “total addressable market.” It’s a maximum theoretical number for the size of the market (not your company’s revenue) under ideal assumptions. A $26 trillion TAM is high but it’s not “unhinged.” For example, the logistics and transportation market is over $10 trillion and expected to double by 2035. Under ideal assumptions, if AI replaces everything from coders to lawyers, why is that “unhinged?”
I also saw a quote from Musk saying that he expects SpaceX to hit $1 trillion in revenue by 2031. Given his track record of predicting performance I think it's safe to ignore such future looking statements from him or companies he controls.
Another way of putting this: global GDP is ~$132Trillion from what I gather.
So this is saying AI products will increase global GDP by about 20%.
The Federal Reserve says AI is contributing about 1% GDP growth per year to the US [0].
So maybe you can get to $13 trillion over a decade just from that. If you assume some acceleration, 20% isn't out of the question.
It is an extremely rosy projection, but if AI can replicate large fractions of the workforce, leaving those humans with the ability to work on other things, it doesn't seem unhinged when you think of it through this lens, just very optimistic - not Elon Musk level optimistic, just "everything goes according to plan and a bunch of things in the causal chain are all slightly on the higher end."
> So this is saying AI products will increase global GDP by about 20%.
No business gets to capture 100% of the value it produces without physical coercion.
For infrastructure that requires high investment, it usually captures something around 5% of it. People tend to work really hard to replace or reduce any kind of infrastructure that gets near 10%. So we are talking about AIs increasing the global GDP by 200% at minimum, 400% more realistically.
With SPCX shares never going down in price, SpaceX can acquire all companies in the US in exchange for its stock, so SpaceX itself is worth at least as much as the US GDP! (/s)
Congrats to the founders, arguably the first true AI-wrapper billionaires? 0 to multigenerational wealth in 4 years is impressive. It is crazy how much more wealth per user that can be created in the AI-space, compared to other products.
Surprising how tech people on a tech forum are some of the biggest Luddites. Maybe it’s because the creative destruction is coming to your industry this time?
This will ultimately be good adding another competitor into the mix with a very strong coding dataset + enough compute to make aggressively top tier models.
This is the card spaceX needs to play to be able to get composer / grok to complete w gpt and Claude
This will be a net positive for our entire ecosystem from a progress and options perspective.
Expensive price but great for cursor shareholders and plenty of demand of spaceX stock at this crazy high valuation.
I don't own spaceX stock at this price
I think Starlink will vastly out perform projection
I think datacenters will under perform
I wonder how many nvidia chips spaceX locked in for next 2 years and I think that numbers is actually the most important number
I stopped using Cursor when I started getting comfortable with Codex/Claude. Cursor is just annoying with the constant popups and it's just not as good. Now my workflow is to use my normal editor, add a todo describing what I want, and then ask Codex+gpt-5.5 to implement it. It absolutely nails it. Using codex is so much more like working with a partner vs the noise and annoyance of Cursor.
That said, I think we're in a narrow window of time right now where any of this matters. Prompt "engineering" and working around your tools will be over in a year or so.
Fwiw I am a c/c++ systems engineer. I think anyone mentioning anecdotal experience like this should clarify. Maybe frontend JavaScript folks have a totally different take and that's expected.
I use cursor 8+ hours/day at work, and have full (and effectively unlimited) access to Claude Code and Codex - tools which I also use personally. I suspect that your "constant popups" were when you were using the editor - a mode that I'll confess I haven't touched in 3+ months.
Workflow in Cursor is actually awesome - I'm a little outdated in how I use it - I still establish goals/objectives, rather than managing the loop which does so - but if you can think broadly enough - I find it's pretty efficient.
Key things I like about Cursor (and I recognize I'm dating myself a bit here): - Plan Mode is really solid - I shift-tab, have it go create the plan using whatever insanely expensive SOTA model is available - I will usually spend 5-10 minutes on the Plan - review it, maybe even tweak it a little. (though 90% of the time it's fine out of the gate)
The nice thing about Cursor (and honestly Claude Code, Codex) - there isn't really any "prompt engineering" involved. You just say, "Go Build me x - it should have y,z features - and build it in golang for me" - and that's it - the 3-4 page Plan comes back - usually pretty credible - and then you click "build.".> there isn't really any "prompt engineering" involved
You should make an experiment; take someone who never used any LLMs or agents, and tell them to use it for the first time in front of you, and tell them to build something like a calculator program or whatnot. Bonus points if they're ICs or at least not-managers.
I think there is a lot us engineers take for granted, when it comes to communicating via text, how to state things clearly and what we think/reason when we read things. A lot of people don't have those "skills" innate, and the first time they use LLMs, they basically don't know how to interact with them, until they realize what they're able to do and not. Then they also learn what to say to steer the model into the right way, this is quite literally a "prompt engineering" skill they're now learning.
You don't even have to go outside engineers. I have teammates that get very little out of Claude Code because the way they integrate their own knowledge doesn't allow them to think of what Claude might not know. They'd say a task was impossible with the tooling, and I'd get instant answers, because I understand what is weird internal business logic sitting 6 repos away, and what is knowledge claude has by default. I can commit Claude.md files for them, but I have to include EVERYTHING, because otherwise they'll let Claude make assumptions and waste minutes, if not hours.
It's a big part of what, in my experience, is separating the very good engineer from the iffy one: Do you have a good mental model, and can you put yourself in the shoes of people sitting in a different mental model? It makes you a better dev, and even more so when it comes to AI tools, which have their own kind of alien brain.
Coding LLMs are distilling developers. It's like the old experiment where you have someone write down the steps to make pancakes and they don't tell you to crack the eggs before adding them to the batter: it takes a particular mindset to be able to make a model of what is supposed to happen and deconstruct that to the level appropriate for implementation.
Until now, the actual act of writing code: terminology, syntax, etc. was a significant hurdle, and that underlying mindset was a very useful, but missing in a surprisingly large number of developers, skill.
Now with LLMs doing the work of "translate this into code," increasingly the only thing that matters is that exact ability. And developers that don't have it or can't develop it won't be developers for long.
or when LLMs won't be able to run on non-existing money any more the scenario will be the opossite.
Thanks for putting into words what I have been seeing a lot at work and haven't been able to put my finger on. We tend to have quite diverse _workflows_ between devs at my company, and success seems to correlate with injecting better context earlier in the process.
I like to chat with Claude about how to approach a given problem, bring in extra context, etc, before even really drafting up a plan, while other people dive into implementation immediately and go on wild goose chases.
90% of the time we end up in the same place in roughly the same amount of time, and there are obviously tradeoffs to spending more time planning vs implementing. I'm oversimplifying as well.
I couldn't agree more. Socratic methodologu, domain modelling, systems thinking, pipes-and-arrows problem solving etc. These are the skills that get real work done in coding agents these days.
This makes a lot of sense and explains why some people are so captivated by modern models, while others see progress as merely incremental.
I'm sure that explains some of it but I really don't think it explains most of the people who have been AI-pilled in the last nine months. There was no amount of context I could give GPT-4o that would make it a net benefit to use that for agentic development. I tried it with quite sophisticated prompt systems and much simpler ones, compendiums of code & business analysis and sparser ones. Yet it just wasted my time - still there were people using Cursor with that model and saying it was life changing. I didn't have that experience until Opus 4.5 - its possible I could have had it earlier but that was when I happened to try it again.
I think many of the people who have become "AI Pilled" (I'll include myself here) had it happen in the last 3 months. Even over the Christmas break, when the Wiggums loop got so much coverage - I still wasn't that blown away going into January/February- 50%+ of the time I'd just write the code myself. I like coding.
But - I don't know if it was April, or May - but very recently - the coding harnesses paired with decent SOTA models like Opus 4.8/GPT 5.5 - just started showing a lot more consistency, and completeness, and sometimes downright clever behavior - that they started to become way more useful.
Just one out of hundred+ examples - I gave Claude Code (Opus 4.8 High) a complex task that involved consul, vault - but I had neglected to give it sandbox permission to download from hashicorp.com. So - it created a entire test harness that simulated both the behavior of Vault and Consul - created all it's test cases, verified that they passed - and when I came back 40 minutes later said that it was all done.
It's test harnesses so accurately simulated the behavior of Vault/Consul - that on first try - no refactoring whatsoever - all of the protobuf/AESGCM/API behavior (that has varied significantly between versions) - worked.
This was something that would have taken me, someone super super familiar with the code and tools and APIs - a minimum of 3 solid days of work - and that would likely involve hundreds of attempts and refactors as I unwound all the weird encryption and packaging layers. It zero-shotted a full solution without having an API to test against
If these agents actually have an actual test-harness - It's honestly hard to imagine what they can't do - subject only to imagination and budget at this point.
Speaking personally - something changed Between January and, Let's say May - in which instead of seeing these things as mostly interesting technology demonstration, in which the flaws outweighed the benefits - I now genuinely think they are the future of programming. I'm dubious that I'll write much software manually in the future - beyond what I do for personal pleasure.
Asked to write a driver for macOS for some thing that didn't have macOS support, GPT-55 found Linux OS firmware on the vendors site, downloaded it, ran binwalk, extracted out the driver, got halfway to reimplementing it on macOS with barely any help from me. I did need to dive into it somewhat to get it across the line, but it showed some ingenuity along the way.
Fantastic post. This sums up my experience perfectly with a near identical time frame to yours.
Which way do you think that goes? Are the ones who "get it" the ones who are captivated or see them as incremental?
I guess all of them?
Some people "got" LLMs back in 2022, others needed it to evolve a bit.
It's not unlike computers. I started using them back in the 90s and absolutely nobody I knew was interested, while today everyone carries one in their pockets...
By that same logic (and I’m agreeing with you as of now), engineers shouldn’t get too comfortable treating “being good at text communication” as a lasting edge. With how quickly agentic coding is evolving, it’s worth considering the possibility that many of the prompting and steering skills we view as valuable today could become far less important in a matter of weeks or months.
Recently I have the SEO guy governing the mostly static, public site with Claude Code. He loves it but you would never imagine the level of mental illness Claude comes up with. If it were an employee I’d literally throw him out the front door, labor laws be damned. And as always, every insane thing it does is some direct echo of its concept and training.
But what's the $60B differentiator here? There are so many similar tools out there. I generally use Opencode, but also Claude code, antigravity and sometimes Kilo code on VS Studio. How can cursor be worth even 10% of 60B?
I don't know what cursors market share is but it feels like 20-25% to me. That is not worth nothing. Then;
1) The data they have flowing through the system that enabled them to build composer (which is much better than stock kimi 2.5) and is presumably allowing the training of a new model on space Xs compute.
2) Cursors new 'github' replacement.
3) Enterprise sales/traction
If you look at all of these together, it's not implausible that they end up mostly 'owning' coding in 5 years time. If they replace GitHub with something more compatible with agentic coding and bring it into their whole ecosystem providing cloud and local agents, PR review and own frontier coding model.
It's specialised vs 'borg' isn't it. One way of thinking is that the world is owned by Anthropic/OpenAI and coding is just one of many things their model and software does. Another view is we have a 'coding with LLMs' company that specialises in this field of endeavour. Hard to say which wins, but I think they have a shot.
Personally my only objection to cursor is that it's more expensive. That's it, otherwise it is great to be able to choose say GPT-5.5 when I want to work on backend and Opus when I want to work on front end. Great to have PR review built in. If they were able to get composer 3 to as good as GPT5.5 / fable at the price of composer 2.5 they'd be winning on price again.
> If you look at all of these together, it's not implausible that they end up mostly 'owning' coding
They really need to change their trajectory then?
And regardless being owned by xAI, a failed AI company which turned into a datacentre operator probably won't help them to achieve that.
> Hard to say which wins, but I think they have a shot.
The market for "coding harnesses" and "AI IDEs" is already oversaturated and they are effectively a commodity at this point, you can use any of them with any provider more or less interchangeably.
> They really need to change their trajectory then? They need to step up progress sure. > And regardless being owned by xAI, a failed AI company which turned into a datacentre operator probably won't help them to achieve that.
I think near unlimited access to compute is exactly what they need to train a frontier level coding model and serve it cheaply and profitably.
> The market for "coding harnesses" and "AI IDEs" is already oversaturated
I think my entire point was that it's not just a AI IDE. It's a coding focused model (currently Composer 2.5, soon hopefully something better), a Github Replacement, PR review/Bug Bot, Cloud Agents and so on and so forth. It's a ecosystem. An enterprise signs a MSA with you and gets everything they need all in one place.
> unlimited access to compute
Yes because Grok failed and they now have "unlimited" compute they can sell to other. I mean you are right that if they did X, Y and Z they could be very successful but their is no indication that might happen. In any meaningfully way seems like Cursor has peaked a while ago.
> An enterprise
Well either they are the type of companies which just buys whatever Microsoft is selling OR they let their developers to mostly pick what they feel is the best tool for the job on their won. I don't think there is that much in between (and its a cutthroat market e.g. GitLab)
> a Github Replacement, PR review/Bug Bot, Cloud Agents
Those things are a dime a dozen, you can vibe code them in weeks/months and there plenty of options on the market already. Well not Github of course, but there are various reason for that which have little to do with product quality and features (not that I think there are many companies which could build a meaningful GH replacement in a realistic time period despite its many flaws).
I just don't really see a huge income stream for dev tools companies (just like there never was) they can skim of something from the top by reselling AI models (generally at zero or negative margins..) but that's not the most lucrative business model when you have no real moot.
How did grok 'fail' ? This is news to me.
By not succeeding? It's an also ran, a closed proprietary model which is behind Anthropic, OpenAI, Google and a a bunch of Chinese companies, how do you make money with a produce like that? (besides the absurd IPO of course...)
At least it didn't succeed yet. They should drop a model somewhere, beating something else in some use case, and maybe people would use.
My company has Claude. People were excited to use Claude. Absolutely no one, despite the option, considered a grok model.
For a lot of people, Grok is the first AI they got to use through Twitter. Grok does get quite a lot of usage, and isn't out of the game - coding tools aren't the only use case for AI.
this is like saying people still use google glass. sure, some people might but AI-wise it is as dead of a product as it gets
Google glass has been discontinued? Besides, many people use it on Twitter everyday. Usage is not limited to what you can see on the Openrouter dashboard.
many people use copilot inside outlook to auto-complete their sentences as well :)
Meaning that Copilot is actually a success, even if you don't like it ;-)
Because Microsoft managed to sell it to a huge number of companies, not directly because people are using it. Hardly anyone is paying for Grok.
US government is paying for Grok to help it send bombs on Iranians. That's a use case.
with this thinking I wish that all your products are as successful as copilot is ;-)
A bad product can be successful with the right distribution. This is what happens with Grok and Copilot.
> Twitter everyday
So what? How much money is that making?
these users are probably losing the company money.
the failure is in converting regular people into actual ai product consumers. Companies are realising that the money is not in regular consumers but in enterprise and they are not considering grok as a serious alternative.
if anything, the name, the branding and the x/twitter affiliation has hurt adoption from money makers rather than help it.
so yes, people know it, but no one is willing to pay for it
Depends, Grok stimulates engagement and pushes to stay on the plaform and feed it data. If anything, it helped justify a massive valuation for SpaceX, which is a metric of success for most corpos.
It helped the valuation but as just like SpaceX hallucinations about the space data centres. Doesn't mean its not a crappy low end model itself. Btw is Twitter even making any money?
"my company doesn't use it so no one uses it" - typical out of touch HN commenter.
Given that Grok is selling all of the compute capacity from its flagship data centre out to a direct competitor sorta speaks for itself.
Does it mean they are out of the race? I have no idea, but things don't look great.
https://news.ycombinator.com/item?id=48037986
They're only selling compute from Colossus 1.
There's a HN article and discussion about Anthropic expanding to use Colossus 2. https://news.ycombinator.com/item?id=48214017 I think it's fairly clear that grok isn't using as much compute as expected.
Seriously though. I haven’t heard anyone use Grok in software engineering context. Maybe I live under a rock.
There are more uses to AI than just software engineering…
So far seems like none of those use cases have generated meaningful income streams? The consumer/non-developer market is mostly dominated by OpenAI and Google anyway...
[dead]
> The market for "coding harnesses" and "AI IDEs" is already oversaturated and they are effectively a commodity at this point, you can use any of them with any provider more or less interchangeably.
Yes and no. I've used a few different harnesses with closed and open models and there is definitely something going on that makes some harnesses work better than others. Many of the differences are hard to pin down and some are things people don't care about. But I wouldn't say they are commodified just yet.
1. Memory use. I have colleagues complaining that Clause Code uses several GB of memory. Meanwhile I haven't heard about that regarding codex or goose, or even opencode for that matter.
2. Suitability for local models. When you use Anthropic models, you use Anthropic as a provider. They can have software between the model and your harness that will fix issues with the model. One notable thing that even the best open weights models struggle with is broken tool calls. There is a lot that a harness can do to fix broken tool calls when working with a straight up ollama running a raw GGUF file.
3. Ease of use with non mainstream models. OpenCode has GREAT coverage of models/providers. Goose, less so as it relies on people to set up their own anthropic or openai compatability settings. e.g. Zed doesn't let you use Z.ai (which, if you speak British English, sounds ironic because "zed ai" isn't directly supported by Zed the editor).
4. Worktree support. Opencode and probably all the TUI harnesses works in a local directory - so you need the terminal to be in the worktree. Zed, however, works centrally on your git repo and tracks the worktrees so you can bounce around your work in a single window.
Of these, '2' is maybe the most important one but also the hardest to pin down as a feature. '3' is a one time cost. Of course '1' could be a blocker for someone using a macbook air or neo.
I agree, Composer Fast 2.5 is getting really good. I started using it for a personal project after I had to switch from Sonnet because I hit the API limits, and I was surprised by how good it has become.
Have you looked at gitlab lately? They have a ton of ai features built in.
I'm not a gitlab user, just learning it, so I can't say how half baked they are or not.
At a high level though it seems like a huge step forward than GitHub
I believe they have some very good training data because of all the data generated by people using the service.
This is the same data they used to finetune Kimi K2.5 to make their newer Composer models, which benchmark substantially better than Kimi K2.5.
I've heard they also want to build their own base models, which will also benefit from their large amount of high-quality training data. Which will solve Grok's model quality problem.
This is all unsourced conjecture of course. But it's what I've heard.
Also from what I understand (not my day job) we're now at the point where the post-training tuning (RLHF etc.) is increasingly important since pre training no longer scales.
So it's not really fair to call it "fine tuning", it's an important part of building a coding model in 2026, and cursor have done a pretty good job with Composer
> How can cursor be worth even 10% of 60B?
Maybe because SpaceX paid with monopoly money (all stock deal)?
It's the data. To do RL.
they are paying for marketshare/customer base. Cursor has a good chunk of it.
xAI overbuilt their data centers - they can't find paying customers for them, that's the reason they made deals with other companies like Google to use their own datacenters.
Cursor has the opposite problem of not having enough capacity. So this works well for them together.
Weather it's worth it - if you beleive that AI will solve every problem then having a piece of the pie early on might be worth it.
Remember how when google bought youtube for 1.65 billions people thought they are crazy? Or when facebook bought instagram.
60B is a crazy number but might be worth it for someone fighting for world dominance :)
you are completely equivocated on most points.
xai is on the line to delivery capacity they already sold to Google and most analysts think they are 50/50 on actually meeting it.
the only proof they have capacity is that musk claims all the money they are burning is going to datacenters and gpu (mostly because if he put it on anything else the lie would be obvious)
Musk is the type of person that would raise billions in funds for a datacenter in space and then just build a datacenter on the ground.
> Remember how when google bought youtube for 1.65 billions people thought they are crazy? Or when facebook bought instagram.
I think these are good examples: in both of those cases the buyer had a plan to monetize.
If you are a user of Cursor, expect to pay more for it or switch.
> they are paying for marketshare/customer base
Or are they paying for talent? It seems like xAI is sorely lacking in talent, most likely due to the CEO and folks' aversion to him. By throwing around some SpaceX monopoly money he can trap some talent with retention clauses and try to invigorate his failed AI business.
I think the argument for Cursor is that it's the dominant tool that enterprises are using for coding, so the theory is Cursor wins that as the "model agnostic", it has a phenomenal Enterprise Sales Team.
From a valuation model - $4B ARR with rapid growth, and the ability to shift traffic to internal models (honestly, massive amount of the time "composer" - their internal model is fine, and obviously going to get better). Say 17x Multiple which isn't unheard for a rapidly growing Startup with solid future structural profit elements (moving to internal model) - that gets you to $68B.
The fact it's agnostic has to be useful.
Being able to compare outcomes for workflows involving competitors will obviously be v v v v useful.
> so the theory is Cursor wins that as the "model agnostic"
But there are many model agnostic harnesses out there: OpenCode, Roo, Cline, and many others. And even Claude Code can be setup to use non-Anthropic models.
As a Cursor user, I don't have to have thought about the providers behind the compute - I get name brand Claude, or cheap Kimi, or Grok, and it's all got roughly the same agentic experience, and only one bill. Enterprises love this.
You get all that at the price of Cursor. Enterprises do love to spend money that's true.
Open routers prices are no different than cursors and you can use any harness you want.
Big brain, small brains? Hmmm
Cursor has BYOK support too. I think it also has Bedrock support.
> And even Claude Code can be setup to use non-Anthropic models.
Too much friction though, with Cursor its out of the box.
Terminal is also model agnostic. Does it matter where you enter your prompt text?
> $4B ARR
If you resell something worth $5 for $5 while having to pay for R&D and operating expenses that's not exactly comparable with a company that's selling actual products.
> Say 17x Multiple
On an extremely low margin business it is, yet again that wouldn't be the stupidest thing in today's market.
>How can cursor be worth even 10% of 60B?
It can't as long as there is plenty of AI without it.
The real differentiatior is that if $60B today turns out to be all thrown away in a worst-case scenario, it would be easily more affordable and there would be less negative impact than $47B at the time if it was all thrown away on Twitter.
Their revenue is 3B, and 20x is pretty typical.
We’re in the new era where startups boast about and bought based on revenue and not on just a number of users with unclear path to monetizing as it had been for the previous couple decades.
We can also note that we see Thrive Capital (Kushner) again in a win.
Where else are you going to get access to a real-time fresh high quality stream of human intelligence to grow your baby AGI? You can’t buy Codex, Claude, Copilot, so what’s left?
Chinese transfer stations?
> Chinese transfer stations?
For anyone that doesn't get the reference, please start here [1].
1: https://www.chinatalk.media/p/how-to-buy-cheap-claude-tokens...
How are you switching between like 5 different editors lol. Bro sloppers will do anything to get their fix. Like the old people at the casino switching slot machines all day based on some occulted understanding that only they think they have.
There is most certainly still prompt engineering involved. How there can be both the responsivity to different cues like "plan this", "write this", "analyze this", "defend this", "poke holes in this", but not responsivity to the various terminology you provide in your explanations of "this", where to get information about specs/standards/requirements, what details I care about, and therefore can't compromise on, vs what details I'm willing to accept whatever the top reddit post from 4 years ago recommends.
I don't see how these systems can have the ability to be effectively expressive about all of the minutia, and not have all of the various different possible expressions lead to vastly different outcomes.
I think all of the cues that you just described are in the plan.
For example - I might (real world example from this morning):
"Create a script that installs hashicorp vault and consul, store the data on consul. Then create ahelper script that will fill the vault server with sample data. Add HTTPS support. Now write a framework that reads and decrypts the encrypted data in consul. Support old (pre 1.3) and new (post 1.3 vault). "
That generates a 6 page plan using Opus 4.8 w/1mm context, including notes on what to prioritize, what format to create the scripts in, etc... (My cursor guidance already has a couple months of hints as to what I want in terms of scaffolding unit tests, canonical linux, performance, security, etc...)
That 6 page plan is the "Prompt" - but it's entirely generated by Cursor/Opus. It's there to tweak if you want to emphasize, or provide some taste - but, honestly - it probably does a better job than I would - so ~90% of the time I just accept the plan as is.
I would say prompt engineering, in the sense of people claiming you need to include in every prompt magic incantations like "You are a senior engineer from a superintelligent alien species" and "take a deep breath and make no mistakes" doesn’t really do that much for everyday work I feel or they are all already included in the system prompt maybe. I reckon it can still edge out a few percentage points in automation.
What actually matters is the ability to communicate well in general, not anything LLM-specific. Being able to state what you want clearly and unambiguously, and having a sense for what additional information you need to dump, even when the other side claims they already have everything they need.
> You just say, "Go Build me x - it should have y,z features - and build it in golang for me" - and that's it - the 3-4 page Plan comes back - usually pretty credible - and then you click "build.".
What you're describing seems like a workflow for building toys only. There's currently no reality in which someone would actually know what the y,z features are before making them. A plan generated in 5min would likely suggest a suboptimal solution compared to what a good solution would look like (which might take a year or two to figure out, for a human, so still a week or so for SOTA models if at all possible). Building something in golang is cute, but hard to be convinced until more novel applications are being generated from prompts.
The data submitted by Cursor's users tho, that seems to be very valuable.
Yes, I tried to use Cursor as an editor. Terrible idea in hindsight.
So your workflow now looks like mine except I prefer a different editor and only use the latest and greatest model so Cursor basically offers nothing over Codex.
I disagree about prompt engineering, but it's one of those things that probably varies because of what language you use, what problems you solve, and the degree to which you care about the output. Unless I'm writing tests, I keep AI on a very short leash because I'm writing critical code used by a very large number of users. I have noticed big differences in output quality depending on how I steer AI. Without steering, it will happily leave in dead code, change the use of variables so they need to be renamed, assume or fail to assume invariants, etc. As I said in another comment, I think we won't need to do that for very much longer, but right now it seems essential.
But that sounds like the same workflow as Codex or Claude, except Cursor is only a harness without its own model? (Or do they have their own model?)
You nailed it - in fact, most of Anthropic's early revenue came from Cursor - much of claude code programming components is essentially a feature copy of Cursor, so it makes sense they are similar.
Cursor does have it's own model - it's a heavily reworked version of KimiK2, called "composer" - that I use a lot of the time when I have fairly straightforward tasks that don't require a lot of exploration or independent thought. Lot cheaper - the Input/CacheWrite/CacheRead/Output costs of Opus 4.8 are $5/$6.25/$0.5/$25 per mm tokens, vs $0.5/-/$0.2/$2.5.
> Key things I like about Cursor (and I recognize I'm dating myself a bit here)
What a world we live in - "dating oneself" is measured in weeks/months! :)
Not trying to be funny but seriously, if these tools can produce a tested 'product' in 45m, shouldn't we be seeing millions of them out there? I mean how far are we from a fully AI built Oracle ERP or even a notepad or helix?
It's a solid question - and to some degree what https://programbench.com/ tries to measure.
Some of the issues (off the top of my head):
- Note - that my "product" was about 3,000 lines of code - so tiny. But https://metr.org/ should give you some insight into the complexity the models are capable of.
- you have to be able to imagine the product. If I have the time, and energy, to imagine what I want - the model will build it. Here is an example of a much better programmer than I and something he wanted built - https://www.boatbomber.com/blog/claude-fable-5
- These are the first drafts. On average - any complex system needs about 10 years and at least 1000 active and enthusiastic about reporting users to really get robust code. Writing if via LLM doesn't (at least so far in my experience) help that much in reducing bugs if you were previously following any semblance of TDD. Lots of bugs in the code - the products you listed above have literally tens of millions of years of user experiences and bug reports that got them to where they are today. No silver bullet yet - just faster, less effort - and it enables non-technical people to create (still buggy) products.
Millions of produced verified software engineered products in 45 minutes in the likeness of Oracle ERP or notepad++, helix are small potatoes when you see the unbounded ambitions of SpaceX in full.
The end point may squeeze quality of operations at the subminute time span for ground control environment seriously launching Starship rockets one an hour, for example.
I think I do this with Claude every day. I don’t see why I need to pay for cursor to get this too.
You absolutely don't. I use all three products. My preference is Claude Code for my personal project. The one at work is kind of sandboxed off - but does have the benefit of an MCP for every enterprise service we have (Kibana, Victoria Metrics, Grafana, Jira, etc...) - which is nice.
Over time - I expect Composer will be cheaper than Opus 4.8 - but the nice thing about Cursor - you can flick between models.
And (this is purely a personal thing) - I really like the extensive collection of "Plans" that cursor tracks - there isn't really a similar thing in Claude Code - but I really like the Claude.AI interface for everything else. It's also a much better general knowledge agent - the Cursor Chat interface isn't as nice.
I’m not sure what you’re on about. I had Claude doing swarm engineering using different models. It would write specs that haiku would implement, it would check itself etc etc. with a simple phrase it goes into planning, multi agent mode, and chews on a problem until it’s done. It’s pretty autonomous.
Maybe you haven’t looked deeper into what modern Claude can do?
The Different Model approach is where from tasks to task - I can switch from Opus 4.8, GPT 5.5 and (very often) composer 2 at 1/10th the cost.
It's not perfect, btw - to some degree you are at the mercy of which models they support - currently only 27 from Gemini, OpenAI, Anthropic, Grok, and Kimi (Just K2.5) - presumably because they have commercial arrangements with them. The "Bring your own Model" model requires you plunk in your API key - which sucks. And only one at a time.
To the best of my Knowledge, Claude Code only supports one model at a time if it's not one from Anthropic (which will use the the entire suite of Anthropic Models depending on the task) - and you have to override it to a single model with an environment variable at startup - no ability to flick between models from task to task.
Depending on your workflow - you can save 70-90% on costs just by chosing a reasonable model for really extensive tasks that don't require thinking, max context, etc....
Different models aren’t subagents - they’re completely orthogonal. I use Gemini subagents for code review in cursor, but mostly use gpt for actual coding.
Same.
When I first used Cursor, I hadn't used any of the "Vibe Code" tools out there, so it was pretty neat to have an assistant directly tied to the editor.
Once I learned how to use Codex, I just used a tmux split with NeoVim and have the effect I wanted. I haven't felt compelled to use Cursor at work since.
I also work with C++, and I use Codex (desktop) which writes 99.99% of my code, plus Visual Studio, which is nice for reading and navigating code. For webdev I do VSCode + Codex.
I started with Cursor back in the day, but switched to Claude Code and then Codex when Cursor got too expensive.
If price wasn't an issue, maybe I'd prefer Cursor only because I can easily switch between models. But that's it. I always disliked the "accept/reject" workflow in cursor, but that's probably optional nowadays I guess?
I love the accept reject flow because I still constantly have to stop AI models from writing awful architecture or reimplementing code we already wrote elsewhere
Yeah, I have found the same. A lot of times it does get things right, but if it deviates man it can just drift hard.
For example, sometimes Claude just obsessively reads files and goes on massive tangents. Then when I stop it and ask, "why are you doing that?", it kindly apologizes and admits it shouldn't have gone on a tangent.
The token burn if I don't stop it would be quite high.
Granted, this might be because I'm not giving it optimal prompt/negative-prompt instructions though.
I just check the git diff after claude code writes stuff. Stage things before letting it run wild so I can undo whatevs.
That's expensive though. The sooner you stop it from acting out the less you spend on a rabbit trail.
How is it different from Keep / Discard in other tools? I've been slowly converting my git repositories to jj locally because that gives me more granular fallback and mix and match options.
Well I tried CLaude Code for the first time in a while (I am building my own coding app www.propelcode.app so I can code on my phone when I take my kids to classes and such) and it literally ignored my question and suggestion and just kept coding away.
I hate the accept reject flow, because I want a conventional code review workflow where I can write comments on specific lines of code and maybe edit the code myself.
If I reject, then the AI will struggle to modify just the parts I disagree with, if I accept, the AI will tend towards adding code rather than updating the bad code.
At that point copy paste without agentic coding tends to work much better.
Fable makes any IDE AI integration almost entirely unnecessary. Claude one shots pretty much everything, and fixing any small errors is easier when just talking to Claude again.
Anthropic is going to offer better pricing using their agentic harness. Why pay more for less?
An IDE at this point is best as a tool for code review. They need to start building better code review tools.
I can't quite understand the "fixing small errors is easier when just talking to Claude" flow.
I tried having it write some tests today. It got very close to what I want, but picked a stupid set of input values (two fields that look independent that should only be used with related values). I thought about "how do I explain this" and then just went in and fixed it myself.
How is it easier to write "Okay, go back to testBlah and change xxx to yyy" versus clicking on XXX in the IDE and typing YYY by hand? Maybe if you had 500 faulty tests and were forbidden from using search-and-replace for some reason.
It makes sense when code generation is the limiting factor, but I end up with a lot of changes where the actual code delta is smaller than the necessary prompt to convince the bot to produce it.
Try the superpowers plugin, let it write a spec (what do you want?) and a plan (how is it implemented). Then let it implement the plan.
Review each step as much as you care. These things take time so you can just do other stuff while it’s cooking.
With proper isolation of projects you can easily have multiple sessions in parallel. I frequently have 4 to 8 parallel Claude Code sessions, each with whole trees of agents reproducing, speccing, planning, implementing and reviewing things.
For common mistakes, you can make it remember things or rely on reviews.
Some of us are working on things that Claude can't one shot. Like, not even close.
Also https://xcancel.com/mitchellh/status/2066657032938442833#m
I really don't see IDE's going out of fashion anytime soon.
> Claude one shots pretty much everything
What?
> An IDE at this point is best as a tool for code review.
I heard from a friend that most devs building serious stuff still write code. It's shocking but true. (No code review needed.)
the reason to use cursor nowadays isn't the IDE (though it's helpful perhaps once a week), but how it makes running models from multiple providers trivial out of the box. I don't have to juggle keys or drop to a shell tool call, it supports calling out to e.g. gemini in a subagent natively. I have multiple models cross-reviewing plans and diffs as a matter of course.
claude code was seriously annoying with the flickering, maybe it's fixed now, I don't know.
cursor also has a (bad) cli if you need it, it seems it's mostly used to setup remote agents, but it does the job in a pinch.
OpenCode and Pi do those things as well, and without a whole annoying IDE bundled in.
OpenCode is miserable from a security perspective. Well clarification the plans they offer where they bundled in free models that train on your use. You are then left to use an OpenRouter which I find pretty flaky for at least the leading Chinese models.
I doubt most people use OpenCode coding plans, nor do they use OpenRouter. I use subscription plans from ChatGPT, z.ai, MiniMax & Xiaomi with OpenCode. It handles authentication with all of them seamlessly. I switch between models based on task/subtask and based on usage limits. You can get the most value out of a lot of these plans at their second-tier and they are often switching in value relative to each other, so it makes sense to arbitrage them like this.
Most of that switching is automated (oh-my-openagent - defaults sub-tasks to different roles, so for example I use MiniMax for explorer tasks and GPT 5.5 for deep design & review tasks, and GLM 5.2 for general orchestrator & most coding). If I hit usage limits it switches to a backup for that task. I'm not sure Cursor authenticates with all the subscription coding plans from all those companies - but if it does it can't be doing it any better.
I run it in a sandbox and its not phoning home.
Which is cool but I think it’s an important callout because it’s shady how they do it in my opinion.
I just use my ChatGPT subscription with it. Not sure what you mean about security.
“Well clarification the plans they offer where they bundled in free models that train on your use.”
Just what I said. They offer paid plans through their tool. Said paid plans are kind of a dark pattern where it’s not immediately obvious the models are training on your data. The harness is fine but that kind of business turns me off and I am usually pretty neutral about those sorts of things.
For what it's worth, flickering in CC has been fixed since around the beginning of the year.
I still saw a lot of flickering in VS Code (I simply use CC as a terminal in VS Code, without the plugin) as of 2 weeks ago. I think it's a combination of CC bugs + Electron(?) rendering the VS Code uses for terminal.
Moved on to Zed (native Rust rendering) 2 weeks ago -> nothing flickers.
Sadly, with Fable 5 cutoff, I am actively exploring CC alternatives. Pi/OMP.sh works great as an agent (definitely better than CC). GPT is seemingly not as good as Opus, but with better agent and better skills, it probaly won't matter anyway. GPT lets you use any agent on Pro subscription.
Maybe flickering, but it's still broken in various ways. Only a few days ago I had an issue where the text I was typing was outside of the textbox frame. Resizing the terminal still maintained the broken view.
The rendering still breaks many times a day for me, in fairly catastrophic ways. Usually because I have the audacity to resize my terminal window.
Ctrl+c -> new tab -> `claude —resume` is deeply ingrained at this point.
It's curious that the person claiming LLMs will soon skip code entirely and go straight to binary is willing to spend $60bn on Cursor.
I’m sure he has a good reason
Also C++ engineer, but from my perspective, for large tasks, agentic coding is still lacking no matter how well I describe desired output. So in that cases I fall back to manual coding and cursor tab helps a lot with boring parts
Companies with large C/C++ codebases should sell AI companies the right to train on their code for $$.
Define "large tasks".
I actually don't let AI take on large tasks beyond test writing and refactoring helper scripts/utils. I keep it on a very short leash for driver/middleware code since the quality bar needs to be extremely high for our codebase. Up until recently I didn't even trust it for that, but some experiments show it's fairly good and even detected issues outside of the refactored functions which I did let it touch. This is with a good amount of 'thought engineering' though where I try to think hard about how to emphasize certain factors and define the problem as best I can.
I've used AI but not through Cursor to implement a high-performance serialization library using expression templates. Non-trivial and few kLoC targeting a soft-realtime and critical system. Would you consider that a large task?
On the flipside, I enjoy Cursor now and came back to it after leaving it over a year ago. The 2.5 model is fast as hell and very good. And whatever harness they have it's terrific, great results. I also really enjoy the fact that I can open my website in the Cursor in-app browser and just click and reference stuff. It's a really cracked workflow. The models can only get better for them.
I would also add that Cursor's "Debug" harness is incredible. Hit "Tab" in the AI editor to Tab through the options (Plan, Multitask, Ask, etc.)
If you do any kind of on-device work, it will spin up a local HTTP log server, and pipe logs from your real device (phone, hardware, etc.) to the server and do realtime debugging.
Claude will mostly guess, have you copy + paste logs, etc.
> I can open my website in the Cursor in-app browser and just click and reference stuff.
I’ve never used cursor and have only seen it in a couple work lunch and learn demos. I’ve never seen that feature. I have a lot of use cases where I’m asking cc to move a widget down a little bit or make a data table full width etc. Being able to reference the actual UI would be useful.
Dunno, Cursor's agents are now more-less equal to Claude Code, just the workflow is slightly different. I like the IDE integration for some projects, allowing me to quickly inspect/review/change/search code, while running Claude Code/Codex/OpenCode/Pi/Hermes on different projects often with local models and it's mostly a question about your personal development style instead of inherent tool capabilities.
It is possible to use Cursor via ACP, so you can use it in any editor that supports ACP (notably the JetBrains IDEs). Our company went all in with Cursor and at the same time centrally disabled the AI functionality of JetBrains IDEs, but a pretty large group of developers (me included) were so vocal about wanting to continue to use our "old" IDEs that IT eventually relented and enabled the plugins needed to support Cursor.
You know you can open the same project in cursor so agent does its own stuff and then opens JetBrains IDE to do your code navigation etc. ?
I am pissed off by people calling Cursor an IDE … Cursor is text editor with AI agents bolted on. I still like what agents do and how the context is managed in Cursor but it is far far away from proper IDE.
Cursor also seems to be doing something with the Claude models that makes it way slower and less efficient as times goes by.
Or it could be just Claude CLI doing something very well.
What are you saying is going to be over in a year or so?
Right now I think there is an edge to how you construct prompts and config files. There is a large difference between "modify f() to do..." and "modify f() to do... Review the current variables and make sure they are still used consistent with their naming. Look for unreachable and dead code. Examine callers and called functions for side effects from the introduced changes...".
I don't think that will make much difference in a year.
I'm increasingly convinced of the opposite. IMO Fable was pretty similarly capable for my day to day work as Opus.
I think there's a pretty good chance that we've reached the point of diminishing returns, for our specific use case.
There are still like a billion other (more difficult) use cases to be tackled, but I think "generating code" has gotten really good to the point where the other bottlenecks will prevent further exponential progress on this specific task.
That's not going away.
It's already going away for me in a sense as I build up a library of AGENTS.md and Codex skills. I see no reason such things won't get baked in at the agent layer so that domain specific rules and such are automatically applied when appopriate.
I'm not sure if you're ahead of me or behind me on this curve, but fwiw, my experience has been that we have now encoded everything that is useful in the various markdown files and have reached the point of diminishing returns on this, with more powerful new models making noticeable but not revolutionary improvements as they come out.
You're essentially making the case here that your work is now automated into a set of one-shot actions that can be performed by an AI model and your job has become to selectively apply these actions. That says either a) we don't do the same work, and instead you're doing some kind of low level devops function that I've only ever seen in rare cases where a human isn't needed anymore, or b) you've vastly oversimplified the software engineering you're doing.
Sophisticated chain of reasoning LLMs like ChatGPT have baked in some natural language operations and they make it so i can create at a higher level of the language expression stack. But I'm still formulating my own expression. There's no conceivable path I can see where an improved model is going to be able to do what I do. I think that is clear from my ChatGPT threads at least.
I think you're reading quite a bit into my comment.. I'll try to respond to a more accurate response to my comment but I'm not going to waste time with this sort of response.
Who's to say it won't?
Who's to say it will? :)
Yeah, uh, why would it go away? In what world do you completely surrender your ability to control the work product, the methods for achieving said work product, etc. That is the dream of a PHB.
Not OP, but I generally agree. Models are powerful enough now to reliably instruct other models. They don’t need fancy tools or IDEs, just the command line.
With deterministic workflows, type-safe languages and test suites, agentic loops pretty much “can’t fail”. They will continue until the types resolve, the tests pass, and the project requirements are deterministically met.
By that point it’s literally just a case of typing a prompt in to a text field, and waiting.
"project requirements are deterministically met" makes it sound so easy
This seems true to me in theory, but not in practice.
[dead]
Same path as you. Went from $60 cursor plan (often exceeding it which costed more in API) to a limitless $100 codex plan where I basically say "read the markdown and implement the instructions". Deepseek also works quite well, surprisingly!
(FWIW Im mostly using python for OCR, LLM calls, data analysis..)
I think I'm late to the party with cursor but I don't use it as an editor at all, I keep VS Code open on another screen for that. All I do in there is agent sessions. I would be open to something else but all the comparisons I see are out of date and talk about the IDE a lot.
The comparison is with Claude Code and codex (and open harnesses like opencode and Pi). IMO they are both better, if you aren't interested in the IDE functionality.
I like your take and think the key takeaway is that there is no single answer for everyone. It’s like eMacs vs vim.
My one question is what popups exist in cursor? It is my daily driver and I cannot recall any popups.
The code suggestions. It's highly distracting and pulls me out of my flow. I know how to code and I don't mind typing. I don't need AI making trivial suggestions. I want it to do exactly what I tell it to do.
You can turn that off.
Sure but then why use it? I like my editor. Codex CLI and vim/slickedit works for me.
'why use it?', it has more than one feature, this is just one that most people turn off.
I like cursor, but I'm assuming they're talking about how it hijacks your tab key. It's amazing when it works, and infuriating when I just want to insert a damn tab!
Maybe that is it and agree.
You can disable that.
Composer is fairly decent. Many people aren’t in the market for an IDE — and a subpar one at that —, but they could sell API access to Composer itself.
"Prompt "engineering" and working around your tools will be over in a year or so."
What do you mean by that? What is happening in just over a year or so?
I think Fable gave a bit of a sneak peek into the future.
My objective KPI: for the few days I was using Fable (18hr a day), it would frequently push back against my design ideas and propose alternatives -- and they almost always felt better to me. Back to Opus now, still 18hr days - and I dont think it disagreed with me meaningfully even once since Saturdy. I consider myself and old hand -- and i think Fable really didn't need me to be very specific in my prompts, it would have done a good job regardless, or even despite my prompting.
Of course whether this is the future is anyone's guess. Maybe we will experience a butlerian jihad and there won't be any prompting whatsoever for completely different reasons :-)
Remember to go outside once in a while, my dude
Crunch time, not the norm.
The models are getting better at agentic coding, so over time using complicated harnesses and precise prompt engineering to attempt to squeeze out an extra X% performance will become irrelevant as the models approach expert-level performance. The bitter lesson in miniature.
there will always be a difference between the general capabilities, and the particularities of your exact environment and requirements.
Closing this gap is done in the harness, either through Skills, user behaviour/prompts , Agents.md etc etc.
I think that this is an area worth investing time in, but it is indeed hard to know what the scope of this is.
>> Fwiw I am a c/c++ systems engineer. I think anyone mentioning anecdotal experience like this should clarify. Maybe frontend JavaScript folks have a totally different take and that's expected.
Similar experience, having transitioned from full-stack to a dedicated C/C++ stack, learned/experienced firsthand that there is no one-size-fits-all tool.
multiple organizations I contract with have killed their cursor enterprise plans over the past several weeks
to me, this seems like the perfect time for Cursor to exit and even "Q3 completion" is too late. Deal just needs to close. Fortunately Q3 completion could mean July 1st too
Yep in my experience the weakest engs in my org are the ones still using Cursor. not a good outlook IMO
I know this is not always true. But the same people who like cursor still are the same people who are less familiar with the terminal.
And I don't know what it is but it feels the less familiar you are with a terminal, the less skilled you tend to be.
Definitely not a 100% case. But has been common in my experience
I taught myself assembly language from a book on a 286, I cracked games with SoftICE as a teenager, tried out every Linux distribution in the 90s, and have been developing software professionally for 2 decades. I prefer Cursor.
Am I an outlier or do you just judge people for weird reasons? I’ve never seen an IDE person judge a terminal person, it’s always the other way around - what’s up with that?
Did you win the Putnam? If not, then don't be bolder than this guy
> Definitely not a 100% case.
Never said I was judging, just making an observation. And to answer - yes by book you would be an outlier.
Its just an anecdotal experience.
I think it's more of a sign of a good engineer. I know a number of engineers that are good and don't really work with the terminal. On the other hand, every engineer I've worked with who was a 'terminal guy' was great. I think being good with the terminal is a signal that the person is willing to 'dig in' and understand stuff at a lower level.
> I think being good with the terminal is a signal that the person is willing to 'dig in' and understand stuff at a lower level.
You know what else is a good sign of them willing to dig in and understand stuff at a lower level ? If they dig in and understan stuff at a lower level. Let's judge people on what matters - the actual work and value they bring. Not status symbols like their IDE of choice or how fast they type into the keyboard.
Yes, that's why I said 'signal' and not 'sole metric used to determine worth'. Devs get so touchy about this subject. I didn't say being good with the terminal is all that matters, it's just an extra piece of information.
yeah its not a good piece of information but you do you
[dead]
I’m very comfortable with the terminal, but let’s be honest. It’s not very good at certain apps. For example, copy/pasting long bits of code or strings into and out of Claude code is highly annoying. Line breaks in weird spots, because of the terminal, for example.
Anyways, I use cursor for a number of reasons:
1. I still want very quick access to the code in the editor. So I want the IDE.
2. Generally solid defaults. Auto-compaction, plan mode, etc, all work pretty well.
3. When I switched back to it from Claude code, it was genuinely faster at running Opus than Claude code. Claude code was grinding to a fucking halt every two minutes.
4. So annoying to search and view your chat history in Claude code. I’m a visual person. I also want all my repos loaded into a big workspace. Cursor also does that great out of the box.
5. I don’t have time to redo my terminal setup again to optimize it for Claude.
Tbh, I’m not aware of much that Claude code does that you can’t also do in cursor. At the end of the day, the agent loop and tools are not that different, and the model is identical.
The tool you use to prompt it is not the hard part. I just work faster when I have everything easily accessible in one spot, which was easier for me to accomplish with cursor than Claude. I found it just got out of my way.
Someone instructing AI through the terminal is a bit like an office worker with a tool belt. I don't think you can say anything about their coding ability until they are coding without AI. Even if thats in notepad.
I actually never mentioned anything about actually using the AI tools integrated into Cursor in my post.
I think I'd generalize my post more to say the more often somebody reaches for the terminal, in my anecdotal experience the more proficient they tend to be.
Proficient in what, terminal use? :) Shocking!
Touche )))
Cursor has a terminal based app that’s just as good as any of the other mainstream ones…
Does it have a plugin library as extensive as Codex? I've started to leverage the plugin ecosystem to fuse data from chat history, wikis, emails, etc.
I know, I actually use it pretty often at work. I agree that its pretty on par with the others.
Totally disagree. I find people still using cursor or other IDE centric flows want to review the code and be more interactive. Claude Code and Codex push agent autonomy and speed. Sorry but they go off the rails too much.
I am pissed off by people calling Cursor an IDE.
VSCode it is based on is text editor.
AI features are great but it is not IDE.
This. I hate VSCode as an IDE and is the reason why I have not used Cursor. I wish Jetbrains actually had some brains to build a better coding agent inside their IDEs (which I think are one of the best out there), but for now Im stuck with codex/cc + Jetbrains IDE
What do you define as an IDE then? I think something with an integrated terminal, file system and git UI is the essential set for most development.
IDE ships it's own compiler, debugger, usually some sort of container/container orchestrator and database tool. I.e. the things that separate hobby projects from production ones. VSCode has these things but they're extensions, not build into the editor
I am not trying to be pedantic, I am really just curious because I’ve never heard someone saying something like vscode isn’t an IDE. I am mostly curious because those of us that write languages like Ruby or python with no compiler that typically debug in the terminal or repl feel like we have an IDE and refer to it as such.
I also think your characterization about hobby/production fall short. My fastAPI and rails codebases anre definitely not a hobby.
Thanks for clarifying though. Interesting to hear your perspective.
> I also think your characterization about hobby/production fall short. My fastAPI and rails codebases anre definitely not a hobby.
That's not what I said?
What I said the was the line between hobby/production is use of the tools themselves. Doesn't matter if you use them via a terminal, VSCode extension, or an external app.
Line between editor/IDE is the former requires downloading extensions but the latter bundles them in the final application.
So long as I can do those things within the editor, it is an IDE in every respect other than pedantry.
You're right, the meaning has become pedantic so I'd put forward a new term because the line between editor and IDE is blurred... but there's still a vast difference between editing on vscode vs IDEA or Goland.
ILE? Integrated Language Environment. The former caters to a specific language ecosystem. Text editors are polyglots by contrast.
It is not pedantic.
If you call text editor that has terminal bolted on that’s not IDE. It is exactly the same as if you would open terminal on the side. Grepping/search with regex is also far from code navigation that is available in an IDE.
Debugging tools and code navigation in a proper IDE like Visual Studio or JetBrains Rider is totally different development experience than editor with terminal.
I guess you never got to debug multithreaded applications in Visual Studio or JetBrains Rider.
Debugging and code navigation are much better in proper IDE. In text editors you grep/search with regex.
why do annoying engineers have such a weird fetish/superiority complex about the terminal. Its an inherently inferior UI. Theres absolutely nothing you can do in a terminal that you cant do in a GUI, and every TUI is just jumping through insane hoops to support functionality thats trivial in a GUI.
Why on earth would you want to look at a code diff in a terminal? Why on earth would you want to use weird bespoke keyboard shortcuts to navigate sub-agents in a TUI session. Why on earth would i not want the ability to open/edit a file manually in the tool im using to write code.
can someone please explain this to me?
As someone who has spent the last 10+ years working in Tmux - but is entirely comfortable on Mac, Windows and Linux desktop environments - here are the key reasons why the terminal experience is superior for me.
- I work a lot with data - and streaming data through text tools is twitch fast. If someone has a question about data - before anybody else can log in to their superset, or analytics database, and try and work through the SQL queries or charts to get the answer - I've already jammed the data through awk and got an answer.
- As an SRE - I work with a lot of systems that have pretty rich APIs - so being able to send a request, get the answer back in json, dump it into jq, select the parts I care about - maybe -c to compress it and ripgrep a subset out - is just fast.
- I work in a lot of contexts with a lot of different systems, datacenters, applications - tmux lets me keep all of them cleanly organized in a separate windows and subpanes. I'll have 15-20 windows open per week, and maybe a 5-6 panes in each- keeping 100+ different contexts (and scroll backs, bash history) - all nicely organized is really useful.
- I'm also a systems guy - and there is no other way to dig into a system but the terminal - netstat, ps, dmesg, /proc - these are all components that have only one credible path to investigation and discovery. If you aren't super comfortable in the terminal - zero way to learn about this stuff.
- Working remotely - means ssh. So - once again - terminal.
The Focus on the terminal is that it's the best tool (and in some cases the only tool) for so many of these tasks - and by performing these tasks a lot - you learn about systems - so the people who spend a lot of time in the terminal tend to know a lot more about systems than people who don't.
And now there are tools like Warp Terminal which make me feel like all that knowledge about command line tools is just dead weight in my brain.
Oh- this is super intriguing - can you share the highlight elements of Warp Terminal that make it feel so useful to you?
It's a terminal emulator with a command prompt that also gives you access to LLMs. But the interesting bit for people who work on a lot of machines is that this also works when you ssh to any remote machine.
So you can bring your agents to any remote system, it even works somewhat well for network devices.
The thing you can do in the terminal that you can't do in a GUI, is glue together over 50 years of useful tools, no matter where you got them from or if the authors have ever heard of each other.
If your workflow fits entirely within a single app's GUI, then yeah, the terminal version of that app is not going to be as useful. But if that app doesn't exist yet, you can put together an 80% version of it for 20% of the work.
Historically, it's also a lot more resistant to rot. Brian Kernighan isn't going to start charging a subscription fee for AWK - and if he did, there are many forks and similar tools.
And, addressing a specific point - why would I want to view a code diff in a terminal? Sure, 'diff old.txt new.txt' is probably less useful than popping it open in a nice GUI with highlighting. But "diff old.txt new.txt | grep '^+'" will only show me added lines, or "| less" and type "/foobar" to jump to all mentions of foobar.
And this is like, the least you can do - the stuff you learn in the second class of "Using the Terminal 101". You can easily use this with git, as a building block to make a quick script to graph the number of changes over time in your repo. Yes, there's probably a GUI somewhere in your workflow that can show this (maybe you click around in Github to find it). But, maybe you also want to just filter that to changes in a specific module in the codebase, or an author, or quantify what module changed the most each month. If you've learnt the building blocks, the scriptability of the terminal lets you put that together quickly.
Anything you do in the terminal can trivially be scripted (automated). It's a self reinforcing loop of making life easier. After many years of working in the terminal, and making little scripts, my workflow is extremely fast, comfortable, and customized to me. You can do some of this via GUI tools, but terminal makes customization easy. Also, using keyboard shortcuts is just vastly superior to using the mouse, you can't convince me otherwise.
> Why on earth would i not want the ability to open/edit a file manually in the tool im using to write code.
I don't know what you mean by this. You can open any file in neovim at any time without leaving the program.
Being familiar with the terminal also makes building CI for the team trivial because I'm already familiar with how all the commands work in the CLI. I'm basically the goto 'devops' guy because I'm one of the few people who actually knows how to work in a Unix terminal.
I will say, TUI is not the same as CLI. I don't find a meaningful difference between a TUI and a GUI other than being able to use tmux or something for window management. I prefer gui tools for database management, querying, git diffs, email, all kinds of stuff.
As for the superiority complex. I've got no judgement on people who prefer the GUI. I have many excellent coworkers who primarily use GUI tools. Having said that, every engineer I've met that works primarily in the terminal has been great. It's a very strong signal of technical competence in my opinion, but terminal familiarity being a signal of competence, in no way makes GUI usage a signal of incompetence.
Absolutely nothing? That's certainly not true. My experience is that those who grew up learning the command line are so familiar with it that they excel at navigating those bespoke keystrokes more quickly than any GUI user who has to scroll, point, and click. Add to that features like command history and autocomplete, and shell users are often far more productive than GUI users.
Fundamentally you are right, problem is that most UI applications in general have garbage tier UX and/or are a buggy mess.
Like any other computer UI: Terminal-based programs (whether ultimately windowed in a GUI or not) didn't start off being familiar. But for those who use them, they eventually become familiar.
And that familiarity transfers between different systems. Windows, Mac, Linux, whatever: The flow of any particular terminal-based program is the same everywhere that it can be used.
It's tidy, and light. It's also network-transparent, and things like ssh keep it secure. Multi-user support is the norm instead of the exception on systems where terminals are common. It doesn't interrupt anyone else's work like something like using Anydesk to access some GUI desktop somewhere else can.
The keyboard shortcuts are annoying at first, but they're faster than mousing around in a GUI -- and once learned, they're approximately impossible to forget.
(You're free to hate terminals if you wish. I don't care if you justify it; I'm not your boss.)
I did a lot of automation at my last job, which was closer to engineering in the classical sense than engineering in the software sense. The automation mostly amounted to web scraping and interacting with local systems, with a generous amount of logic in between. The largest roadblock wasn't the scraping or the logic, it was the inbuilt assumption of the local programs that there would be an ape with a mouse and keyboard driving the interaction. Outputs that could have been text needed to be copied to and read from the clipboard, or exported as a spreadsheet. Inputs that could have been text were only accepted in spreadsheet form. Pixels needed to be read from the screen to tell when one step was complete and the next could begin. Mouse clicks and keystrokes sent to and fro when it all could have been a series of commands. I cannot count how many written processes and procedures existed could have been a bash simple bash script.
I'm not arguing that these programs should not have a GUI, for that was the simplest way to use them, but the lack of command line functionality places a hard limit on the productivity of an organization, and ensures that the only progress on that front comes at the expensive of exceedingly limited developer time.
"But nobody knows how to use a terminal anymore," I hear you say. Well of course they don't, nobody under 35 without a background in programming has ever had cause to use one. We made everything so simple that nobody ever has to learn anything. That isn't to say that people cannot learn, just they have been robbed of the natural opportunity to do so. Otherwise intelligent people never progress beyond the manual step-by-step interaction that passes for "using a computer" today.
A computer is a tool in much the same way that the a machine shop is a tool: it is a tool that can build other tools. The role of software developers should not just be to build simple tools that can do one task in isolation, it should be to build tools that less technical people can use to build the tools they find themselves in need of. GUI-only programs are simply not fit for that purpose because they lack composability of simpler terminal-based programs.
Different strokes for different folks, but unfortunately they take their opinions and preferences as a sign that others are inferior.
Yah this judgment and arrogance is so annoying in tech. And worse it stops us from learning. Some of the best lessons of my career were when a new developer asked a question often taken for granted or we implemented a design pattern to make coding more approachable.
I think you miss crucial point here.
Terminal/CLI is superior if you know what you want to accomplish. There is no faster way of doing stuff on a computer if you know what you’re doing and having tab completion and knowing all the incantations.
GUI is superior when you have to figure out what to do and what needs to be done. GIT branches state checking, finding out a switch to enable a feature so much easier in graphical interfaces.
I find that what you call „annoying engineers” are just people who have their tools and use them efficiently … but once they have to step out of their path they become annoying or even obnoxious.
The opposite is random person who expects everything to be easy and available in less than 3 clicks — well good luck engineering an interface for complex system that does that, not everything is Facebook, there are systems where you need to spend time wrapping your head around.
I refuse to believe this wasn't written to intentionally bait, reads like copypasta.
Honestly the TUI in most of these coding agents is so fancy I have trouble thinking of them as "terminal". I use Pi Coding Agent and the fact that it's terminal means it's easy to run inside something properly sandboxed in a YOLO mode using normal bash commands instead of relying on individually sandboxed tools.
Once I got the tmux settings for proper scrolling and whatnot it feels fine. Honestly the TUI of tmux is the one that really enrages me - so much complexity for just "I want to switch terminals on my remote".
Zellij is a pretty good tmux alternative, with a UI that feels a lot friendlier.
With the one caveat that it does not (yet) support copy/paste buffers or copying from the scroll back buffer. It’s also adamant that you use a mouse to select/copy - anything more complex and you need to pipe everything into neovim (or whatever text editor you use) and do your work there. I love zellij - it feels like the future - but hard to give up keyboard based select/copy and pulling from my scrollback.
why do annoying engineers has such a weird fetish/superiority complex about the terminal. Its an inherently inferior UI. Theres absolutely nothing you can do in a terminal that you cant do in a GUI, and every TUI is just jumping through insane hoops to support functionality thats trivial in a GUI. And guess what, you can just open a terminal in cursor! who knew!?
Why on earth would you want to look at a code diff in a terminal? Why on earth would you want to use weird bespoke keyboard shortcuts to navigate sub-agents in a TUI session.
can someone please explain this to me?
I am not one of them but quite a few programmers prefer not having to use the mouse at all when working.
The terminal is an old but astonishingly powerful user interface that is still evolving.
Good terminals can be very snappy and configurable in ways that most GUI are not.
There is also arguably an aesthetic/fetishism appeal to it.
I've worked in the terminal at some point of my career, as there was not many other choices, and I understand how someone can get really used to it.
Speed and scriptability
What's a polite way to suggest you ask AI first?
Why should others spend their valuable time helping you? Especially when you insult the people you want to answer you "fetish/superiority complex" just demonstrates your own prejudice.
Personally I ask AI for a summary of positions, and prompt to provide some good articles on a subject - ideally articles from supporters of either side.
id prefer a human to explain it to me
I actually was a Cursor advocate / CC hater (go back in my comment history), and now I use only TUI coding harnesses.
To start a big part is just the efficacy of them, which comes down to the model and the harness logic itself. CC is good, it's sub-agents, loops, background jobs / agents, skills/hooks/etc have typically been pretty far ahead though others are constantly catching up.
But you're sort of missing something. I use iTerm, so to me it's not the TUI itself, it's iTerm. And while it's imperfect, what I get is this:
I can open and close sessions nearly instantly and tile and window and tab them as flexibly as I want, plus it's a system I'm familiar with in terms of shortcuts etc. Has my configured theme, fonts, etc all set up. Every GUI app is different, every TUI app has half of the UI already incredibly familiar to me, it's not "just text", it's iTerm.
That also means they all are the same - I run Codex and Claude and pi side by side, and i switch between them with no overhead and minimal mental model shift. Sure, different harness does suck but that's the same issue with GUI just with an additional new layer to learn.
Smaller thing is because it's all text, there's no limits on my ability to copy things out. And it's a really fast text renderer that can render tens of thousands of rows efficiently. Many GUIs have various dialogs, unselectable areas, virtualization, or just slow past a point. I trust my terminal scales.
Just a few reasons.
Try being more respectful and inclusive with your request first.
I don’t think mapping tooling to ability makes sense here, particularly when the “advanced” tools here just abstract more away, though I agree that Cursor is terrible. So many useless windows.
Opposite. The weakest engineers trust CC or codex, stopped reviewing the code and push slop PRs. Those still acting in the loop move faster with better architecture and coding patterns and aren't losing their skills.
You must be a joy to work with.
I hate to be the one to break it to you but the weakest engineers are going to be producing just as much value
>That said, I think we're in a narrow window of time right now where any of this matters. Prompt "engineering" and working around your tools will be over in a year or so.
How so?
As previously C system programmer, I barely use IDE, cli agent for coding and desktop codex for various non coding tasks. Nowadays settled to codex/gpt5.5 and it does really good job
I didn't like cursor when it first came about but now I use it for my personal projects. The plan is good value for accessing different sota models occasionally for planning. Composer is actually really good, and fast.
I'm not sure if it's because my personal projects are small enough I know them inside out and my work project is huge, but I prefer terminal code agents for day work over ide integration.
i moved all my AI coding over to claude code when claude 4 came out, but kept cursor for the tab complete. since opus 4.5 i haven't really needed the tab complete so i canceled my cursor sub 6 months ago.
switched back to vscode so i'm not exposed to the potential mess that is openvsx too. trying to get used to zed but i'm just so used to vscode
Also using codex as a full time partner these days. What do you think happens in a year or so that changes the way it works around the tools? It becomes the only tool we interact with, and it assumes control over the others?
Could you clarify what you mean by “…will be over in a year or so”. Genuine question. Is it that models will be so good that none of this matters or we will need to go back to older ways?
I recently made an npm package with a small C helper that runs in the background. The JS/TS code is 99.9% unit test covered and for sure "cleaner" code. Just my opinion though.
Cursor is mostly no longer an editor. By default it now opens an agent window only, you have to click a few buttons to actually edit files yourself.
My experience exactly... minus c++
I primarily use Cursor, compared against Claude and haven't used Codex before - to me the benefits:
- Composer 2.5 is cheap, fast, and very effective; I don't even use other models that much anymore, as it's usually marginally better for way more cost, though sometimes I do for specific things like making better translations for our app than a coding-specific model could normally output
- It makes setting up and maintaining Cloud Agents super easy, the agent can basically set up itself and if anything changes that makes it not update properly, then it tells you and can fix itself easily
- Easy to move agents from remotely in the cloud to local and vice versa
- Easy to work with agents via either comments on GitHub, or via the web app on my phone (though it is relatively constrained relative to the actual desktop UI, which is a bummer)
- Code reviews with Bugbot is surprisingly good now vs when it first came out, as is the Security Agent, while being an order of magnitude cheaper than stuff like Claude reviews
- Automations are easy to configure and manage - crons, in response to repo events, etc. For example I don't use Renovate or Dependabot much anymore since LLMs can update deps and investigate subtle breaking changes much better than a dumb version bump script can
- Limits are obvious rather than Anthropic's mysterious quota amount that they don't explain at all
- Queueing up messages rather than the agent taking in new messages mid-work and then trying to mesh them together somehow - I find queueing much more predictable and easier to work with
- Plan mode is good too, but not particularly different than any other agent
- Easy to jump into the Editor view and actually go in and manually code things or interactively code with the LLM when you need to, since sometimes LLMs just suck at doing certain things autonomously. I'm not in the "stop coding bro" camp, I still like to take the wheel fairly often.
Cursors target users are not developers but casual vibe coders.
Literally the opposite is true. Being a text editor at its core, and by spending a lot of effort refining the human-AI pair programming experience, Cursor makes sense for someone who wants to lead code development with AI being a speed multiplier and a team member.
Tools like CC are best suited to vibe coding.
Oh, you should try OpenSpec !
Same.
A space company is buying an IDE for roughly the cost to build 150 of world's most expensive modern hospitals [1]. How is this in SpaceX's interest? Isn't it kinda bizarre that Elon is pivoting SpaceX to something else?
1. https://www.cnbc.com/2026/06/16/spacex-spcx-cursor-acquisiti...
It’s all in a stock that may very well be near its zenith when this closes (or maybe not. This is so far past fundamentals it’s impossible to tell).
They’re spending Monopoly money.
It also seems like SpaceX is poised to Hoover up all of Elons companies so it’s might not be “just a space company” for long.
The company that just IPOed is already overwhelmingly "X AI" financially, regardless of the fact that it says "Space X" in the marketing. Whether SpaceX also buys Tesla is hardly even going to move the needle.
$10 how "space x" will sell for peanuts the starlink portion to musk and keep only the sinking Ai pieces for the bag holders
StarLink is a bad investment too in my opinion. It is fundamentally a US company that must charge US prices but in the US there are relatively few people who need that tech. Only in more rural areas that are sparsely populated do you not have access to fiber or cable which give better speeds and latency.
The majority of their revenue is overseas. But there they can’t reasonably charge $99/month. And there labor and land are cheap such that you can get fiber laid cheaply and quickly.
StarLink is the only thing about SpaceX that makes a profit. But that's only possible because no one else has a network up there hence it can ask for any price it want. Let's not forget that before StarLink there was Iridium and it got ousted because there was a better offer and I wouldn't be surprised if some other company will eventually go up with a better offer.
Uh, all of SpaceX’s launch business has been insanely profitable since 2010. They basically own the entire global launch market with huge margins.
They aren't in the black at all, let alone "insanely" so. Maybe if they just did commercial launches on falcon 9, but instead they've been shoveling those "huge margins" straight back into r&d on starship etc.
The Falcon launch business is profitable on its own, which may be what GP meant. Once you include the ~$3 billion in Starship yearly R&D the space segment goes into the red.
It is what I meant. SpaceX has strong division between its operational and R&D arms, and Falcon launch has been insanely profitable from the first F9, even before reuse dropped their costs like a rock. They just invest those operating profits back into the company.
So they’re playing the long game, building a moat, creating lasting value with the cash surplus, just like Jeff Bezos did.
If anything this should make you more bullish.
Starlink guided drones are tipping the scales for Ukraine right now. That's what enables the range required to successfully hit trains and fuel trucks supplying Crimea through the "land bridge" made from conquered lands in South Ukraine.
That plus satellite comms. Both Russia and Ukraine were very fond of how it improves communication.
Edit: meant to say that in the time of global instability, weapons tech is going to be valuable.
This is putting the world at risk of Kessler Syndrome.
Not so much Ukraine using drones that use starlink, more the sheer number of starlink sats up and planned, the numbers in the proposed other constellations going up and planned, and the ongoing propensity of Chinese rockets to fragment after use rather than safely de orbit.
For interest:
An Orbital House of Cards: Frequent Megaconstellation Close Conjunctions (2025)
~ https://arxiv.org/abs/2512.09643Yep, it feels like forever a niche market. As soon as you think about it like fundamental infrastructure (which as I understand it is what we’re supposed to think about) it quickly becomes obvious that cables are better suited. Wealthy enclaves, digital nomads etc will pay for it, but that’s not going to get the revenue they’re hoping for.
You mention that cables are better suited... but that is the whole point of Starlink. Getting cables everywhere is either not feasible or expensive. The rural properties are already mentioned as a very big source of usage right now around the world. But you also have everyone who is using it for some mobile purpose. Whether that be planes, boats, RV's, cars, military, etc... I've even seen people live streaming things using Starlink due to cell towers not giving good enough connections where they are.
While they may not be ideal or people living in Urban areas, they also aren't limited to selling to people who are physically connected by cable like a regular ISP would be.
I think you're underestimating the market here. Especially on the enterprise side when you start thinking of things like airliners, cruise ships, etc...
Agreed, starlink is obviously advantageous from the global (and extra-global) perspective. Running cables is expensive. But so is launching satellites, and it’s hard to see how you reach a tipping point where satellites gets cheaper than the standard boring cable networking that supports the terrestrial internet
> But so is launching satellites
Starship is trying to get launch cost down to $10 per kilogram to low earth orbit. At that cost it is not clear to me that cables are better except for short distance point to point backbones. One other issue with cables is regulation or monopoly.
There are two different questions. Are cables cheaper and is Starlink profitable in the long term?
Launching a starlink satellite costs $1.5-2M per satellite for the current design. Each satellite lasts around 5 years and they need 42k of them.
Do the math and you get $13B to $17B per year ignoring inflation and ignoring the massive ground infrastructure that is also needed (costing billions more per year).
Residential plans range from $55-175/month and will make up almost all subscriptions. Most people will get the $85 middle plan, but we'll round up to $100/mo or $1200/yr.
Breaking even with no profit requires 10-15 million subscribers.
As starlink claims to have around 12k subscribers and just 10.5k satellites, I'd guess the system is profitable.
The cable question is still compelling though. Rural fiber costs around $25k per mile. This means that an average $15B/yr buys 600,000 miles of installed fiber every single year.
My conclusion is that the ideal solution would be Starlink keeping total satellites closer to the current 10k number ($4-5B per year) and spending the other $10B/yr running fiber to new areas.
Their subscriber base would drop as new fiber got installed in rural areas overlooked as "not profitable enough", but stabilize at still profitable enough from travelers and truly inaccessible places.
There's a limit to orbit at your desired location similar to what real estate is. At some point it becomes scarse and consolidation will happen.
At a time when state actors are becoming increasingly non-cooperative, it may not be wise to put all your eggs in someone else's basket.
When cables not only become expensive but also risky, in an increasingly tense world.
How many people actually live in areas that do not have some sort of broadband? That’s the whole point: this is a product designed for relatively few people.
I think there's a strong chance this is a case of it creating a whole new market though.
There are people whose current behaviour/situations will happen to benefit from this, and that may be a niche, but seems like there's a really solid chance many more people actually will change their behaviour in response to this being available. That's how disruption happens.
To be honest I can easily see the default changing if the service is good enough. I mean it seems like you basically get most of what is good about wired plus a whole load of extra previously totally unavailable benefits. For a price, to be sure, but that'll come down.
> labor and land are cheap such that you can get fiber laid cheaply and quickly.
Which is why all of sub-saharan Africa is crisscrossed in fiber, and every hut in those villages has an ONT zip-tied to the straw walls where they get their symmetrical gigabit service.
What a terrible take.
There is a huge market even just in the US for Starlink. The worldwide market for people who need internet access in remote places is positively gigantic.
Additionally, Starlink could be hugely profitable if they exclusively sold access to ships and aircraft. You'll never be able to run fiber to those.
They can indeed reasonably charge $99/month to many many millions worldwide who don't have any options for low latency, high speed internet access. You vastly underestimate the need for the service.
On the one hand you have a technology product that’s only relevant to rural consumers. Nine out of ten people have cable already.
On the other hand your margins are amazing because all you do is fly little boxes over everyone’s heads launched with government subsidized rockets. No linemen or plant-hire or contractors to sap your profits.
The biggest threat would be commoditized terrestrial wifi / 5G. The more cell service competition there is, the smaller the market for satellite, until it’s only applicable to 1% of the population (and the poorest 1% at that.)
> Nine out of ten people have cable already.
More like 8.2 out of ten. Either way the remainder is still a pretty decent market. And that's just talking about people in the USA. About 75% of Starlink subscribers are outside the USA.
> and the poorest 1% at that
Not by a lot. People who live in remote areas in the USA tend to have much less money overall, but they tend to spend much less money overall, leading to a similar amount of buying power. Someone who lives remote is more likely to own their home outright or have a relatively small mortgage. Their socio-economic status can appear numerically depressed because the numbers generally don't account for non-monetary consumption. (You got paid a salary and bought salmon from the supermarket. Remote dude fishes for salmon in a local stream. You both traded your time for salmon, but remote dude's salmon is invisible to GDP statistics.)
And furthermore, for them, Starlink would be budgeted for like an essential service rather than a luxury convenience.
> Starlink would be budgeted for like an essential service
As someone who has lived and worked remotely for the bulk of time since 1960 or so, it's not essential ipso facto; myself and most of the people I know have somehow managed to survive sans this supposedly essential service for 60 odd years (since 1935 in my fathers case, he's not dead yet).
Its more compelling use case is a relatively cheap way to integrate vehicle GIS data across four to ten thousand hectares or so (ten thousand to twenty five thousand acres) for farming, mining, exploration, etc.
Globally, its not especially attractive for non civil applications (military use, etc) as it creates a reliance that can have a plug pulled at the worst moments.
You are correct that it's not something any military can/should rely on for any future conflicts.
That doesn't make it useless though. Ukraine certainly finds Starlink attractive for military use. Despite all the misleading headlines and de-contextualised quotes, SpaceX has been reliably on Ukraine's side of the conflict and has been an essential communications fabric for both military and civilian.
Just throwing it out there, Im a Verizon customer in the Tri-State Area and I frequently lose service on the train or bus on my way to the city. It's more than just a 'rural issue'. I looked into getting a StarLink plan for my commute until I realized logistically, at best, I'm a freak sitting next to the window with a satellite dish on the train.
No it’s not.
I flew from Zürich to Bangalore via Qatar and both flights had starlink.
There are many many uses for it, besides rural homes.
Yeah i'm no fan of Elon but starlink is actually quite useful. It solves a real problem - I need network access anywhere on the globe, for any reason, at any time. they can easily charge what they wish for that (military, air travel, desert tourists etc.)
> Nine out of ten people have cable already.
What on Earth are you talking about?!? Half the people on Earth don’t have any sort of internet access at all.
> government subsidized rockets
The Starlink launches are not subsidized in any way. Now it’s clear you are either totally uninformed, or actively trolling.
> The biggest threat would be commoditized terrestrial wifi / 5G. The more cell service competition there is, the smaller the market for satellite
This is actually one of their biggest market opportunities. How do you think the backhaul for 5G towers works in extremely remote locations?
Like I said, terrible take.
Most of the areas without good internet barely take home $100/m and a lot of them take home less than that.
If you hooked up every single cargo ship on the planet to starlink, you'd only add around 100k connections and average wages on most of those vessels is $5-8/hr (very few US/EU sailors these days) for a handful of people which tells you how much businesses actually care about their workers who do dangerous jobs.
Elon's award is tied to growing Tesla's market cap - it's pretty transparent that he's just trying to ball-of-mud together everything he can to hit that target and grab the bag.
What would it mean if SpaceX buys Tesla though? Does the combined market cap count? That would be wrong. Tesla buying SpaceX just for hist bonus and then rebranding to X would be classic Musk.
It's a game for him, but so ridiculous. While Tesla was pushing electrification and SpaceX pushing rapid rocket re-use I kind of tolerated Elon's antics, but since he got involved in politics and DOGE I can't bear it anymore.
>What would it mean if SpaceX buys Tesla though? Does the combined market cap count? That would be wrong.
I took a look at the proxy statement as they have it outlined in this [0] document (Proposal 4). As currently formulated, Musk has 12 operational goals (like ship 1 million robots or hit 50 billion EBITDA) and 12 market cap goals. These need to be paired together for the shares to vest; so, if he reaches the first market cap number, he also needs to fulfill an operating goal for the first set of shares to be earned. These earned shares vest in ~5 years.
This comes with a huge caveat. If Tesla changes control, those operational goals go out the window and his stock award is based solely on market cap (along with the shares immediately vesting). The share price for this is the greater of:
1. The last traded Tesla price prior to the acquisition, or 2. The per share price outlined in the acquisition.
The "easiest" way to take advantage is to IPO SpaceX (which he did), pump up the market cap, acquire Tesla for a sizable premium, and vest as much of the stock award as you can. It means you get to avoid the operational goals entirely and vest a bunch of Tesla (but soon to be SpaceX) stock.
[0]: https://www.sec.gov/Archives/edgar/data/1318605/000110465925...
Yeah, that seems like his goal.
> If Tesla changes control
Does selling Tesla to SpaceX actually satisfy "changing control" if he remains the largest shareholder?
My understanding is that Musk doesn't have a controlling share of Tesla. So by the letter of the law, a SpaceX acquisition would qualify as a change in control (because we've gone to a new majority owner).
Conflict of interest laws don’t apply to fascists.
[flagged]
The thing is, even the center of the road academics who refrain from such comparisons are noting the similarity. The MAGA movement meets the traditional defining characteristics of fascio. We can be rigid and say that only Mussolini‘s political movement can be properly be called fascist. But if we accept calling Hitler and the Nazis fascists as well, then we open the door to any movement that meets the criteria. And Trump does fit the bill.
https://substack.com/@rutgerbregman/p-197211597
You are conflating a strong regime with authoritarian tendencies with fascism. The link you gave could be used to define China and India as "Fascist" as well. Which they aren't. Besides, other categories exist to define regimes, but I guess now the public debate is binary: either you are fascist or you aren't. I agree that it reduces the amount of thinking required, which is always nice I guess.
In the case of Trump, caesarism is likely a better term. Unlike fascists leader such as Hitler or Mussolini, Trump doesn't have an ideology, is heavily corrupt and self-serving and is deeply influenced by a foreign power and lobby. Interestingly, he was stabbed in the back by his closest collaborators at the end of his first term, when he tried his luck. Like Caesar.
But Caesarism is less frightening that "fascism", so...let's ignore it! This strategy is a classic of the communist way of doing propaganda: Russians still whine about "Ukrainian fascists", and in the 80´s in the communist block, everyday you could hear complains about "fascists" at the radio. The Berlin wall is originally named the "Anti-Fascist Protection Rampart".
Why are you reusing soviet talking points, exactly?
Caesarism/kleptocracy is definitely a more accurate description.
People usually attribute the term fascist to this admin, because of the racial/ethnic retaliatory actions or the neonazi/neofascist base they need to overpower/outweigh the more moderate right.
Actually, Caesar was very self-interested and got very rich from his campaign in Gaul. Plutarch estimated that Caesar brought back (and sold) a million slaves.
Yes of course, but that's only the admin.
The strong minority base that has been artificially inflated to advance the personal interests of the leadership has linear roots dating back to the XX century rhetoric promoters and activists. In Europe there's the militant and militia groups of Gladio and the likes that thrived in illegal affairs and that have persisted and were integrated into power following 90s (case in point in Italy Berlusconi basically legalizing the MSI, AN, FdI and Lega by using his privately owned televisions and media, to acquire an election victory).
Italy is a good example why it's foolish to use "fascist" to describe everything one doesn't like. The European politically correct press was wary of the new "post-fascist" leader of Italy, promising us black shirts and leather boots again. In reality, Italians got a classic right-wing government, which increased work visas for non-european migrants (as they usually do, since doing the opposite of what you were elected for is a common habit in Europe).
In reality "fascist" is a qualifier to frighten white people voting for a right-wing government. Trump base by the way is not fascist, it's the evangelical bible belt. Which is crumbling at the moment, by the way.
I assume that you read the link I provided. It's structured as a 10-point list. The 10 points of comparison are:
1. a mythic past and national rebirth 2. victimhood and humiliation 3. hierarchy and dehumanization 4. contempt for weakness 5. the cult of action 6. the leader as savior 7. the purification of institutions 8. propaganda and the assault on truth 9. the merger of state and corporate power 10. violence and terror
It's not stretching things in the slightest to say that MAGA and Trump fit these to a T, certainly post-1/6 if not before. If it walks like a duck, and quacks like a duck, we shouldn't be too timid about using the word "duck." Do China under Xi and India under Modi also fit the bill? Maybe. I don't think they check 10/10 boxes perfectly though.
I know we're running up against HN rules, so I'm trying to keep this meta and not object-level. I'm not debating whether any of this is defensible or good or bad policy or whatever. Just whether we should use the term.
The first Trump administration was arguably Caesarism. The second has proven to be an entirely different character. Just picking on one thread that is more on-topic for HN: the merger of state and corporate power (#9 on the list above) is a defining economic characteristic of fascism that we are only now seeing this term, with direct acquisition of large stakes in Intel, IBM, GlobalFoundaries, various rare-earth metals companies, and Westinghouse. OpenAI is in negotiations to do this, and even before concluding a deal the government has already been weaponized against their chief competitor, Anthropic.
Other things have progressed (regressed?) significantly in this second term against the above metrics. Unlike the first term, which mostly rode out the instability, we saw in 2025 far reaching restructuring of the civil service. And trans oppression and scapegoating, for example, is seriously reaching levels comparable to Nazi jew hatred and victimhood. As someone who lost my European relatives to the holocaust, I do not make that comparison lightly.
Seeing as ICE is currently building / converting hundreds of new detention centers all over the country, it is reasonable to be worried.
The definition you are using is far too broad, so you get a barnum effect where every authoritarian political movement could be shoehorned as "fascist". Ask real fascists what they think of Kash Patel.
Benito Mussolini, who first used the term fasci in its modern meaning to describe his party and movement, was not a racist. A cultural nationalist, yes, but he thought the nazi biological race ideology was bullshit and openly said so. The racial discrimination laws were only enacted after he went to Munich, hat-in-hand.
Yeah, anyone who openly calls themselves a fascist are probably white supremacist nazis, and I presume would take a dim view of Kash Patel for that reason alone. But genetics-driven racism (as opposed to cultural nationalism) was never a core tenant of fascism. Fascism requires an us-vs-them othering world view, but it doesn't have to be race-based.
Us-vs-them worldview is I think a very common thing in the world. Was the Soviet Union fascist? Is India fascist due to its extreme antagonization of Pakistan (and vis-versa).
What you say about Mussolini is false by the way and Italy enforced strict racial discrimination in Ethiopia. Political racism was commonplace everywhere at the time and is not very specific to Fascism. Unless 1930´s USA was a fascist country, which I doubt.
https://www.rachelmaddow.com/prequel-by-rachel-maddow/
Rachel Maddow’s ULTRA and PREQUEL podcast series may change your mind on that last claim.
The existence of a fascist movement in the country doesn't make it fascist. France had a similar movement at the same time and wasn't fascist, for example.
It was a few years later.
Let’s not talk about fascism until we’re already fucked? Is that your strategy?
[dead]
Fascists don't support free speech. Please, learn what that word means[0] and then use it responsibly. We're already seeing the effects of misusing and overusing several other important words for political/societal discourse, and this is one of them. It's entirely unhelpful and lowers the quality of conversation here.
[0] The Doctrine of Fascism is available, for free, in several places. Start there.
Does Elon _really_ support free speech?
* He suspended multiple journalists from Twitter/X in December 2022 after they had reported on or criticized him [0]
* He has been accused of retaliating against critics and employees through threats, lawsuits, or firings, rather than tolerating dissent. [1]
* He selective enforcing Twitter/X platform rules. Here's 10 examples: [2]
He tolerates speech he likes and often punishing speech he dislikes. That's not being a "free speech" absolutist.
[0] https://truthout.org/articles/free-speech-watchdogs-condemn-...
[1] https://www.businessinsider.com/free-speech-absolutist-elon-...
[2] https://gizmodo.com/10-times-elon-musk-censored-twitter-user...
> That's not being a "free speech" absolutist.
One doesn't need to be a free speech absolutist to not be a fascist, far from it; and the exceptions prove the rule - those examples are the definition of exceptional. "10 times…" I can remember people getting banned en masse for changing their profile pic to an NPC character. The difference is night and day.
Elon fits the definition in Britannica pretty neatly
[flagged]
- Opposition to Marxism and socialism
I'd agree.
- Opposition to parliamentary democracy
Yes. buying elections
- Opposition to political and cultural liberalism
Yes, banned speech on twitter that trans people use.
- Totalitarian ambitions
Yes, buying elections
- Corporatism
trillionare
- Imperialism
Literally wants to colonize a planet
- Military values
Building a military satelite constellation
- Volksgemeinschaft
White supremacist
- Mass mobilization
election lottery, trump rallys
- The leadership principle
idk what this is
- The “new man”
or this
- Glorification of youth
i guess not
- Education as character building
what
- Decadence and spirituality
No.
- Violence
No.
- Extreme nationalism
xenophobe
- Scapegoating
those darn regulations are always in the way
- Populism
Possibly?
i'm sure your reply will be pedantic
> i'm sure your reply will be pedantic
Have you heard of the term "bad faith argument"? That's a good example.
> - Opposition to political and cultural liberalism
> Yes, banned speech on twitter that trans people use.
I had to check this one. I found two stories. [1] shows that people were warned about speech (e.g. using the term cisgender) but were still able to post without problems. [2] showed that X's rules were updated to prevent harassment in the form of deadnaming or misgendering. I doubt that either supports your assertion to be true.
As to the rest of your answers, you seem to have given away that you didn't read the Britannica article at all:
> - Education as character building
> what
You should know, right? You read the article and made the accusation. You've done this several times:
> - The leadership principle
> idk what this is
> - The “new man”
> or this
There's not really a need for me to be pedantic in the face of such slipshod work, but then again, perhaps you think that not being slipshod is being pedantic. Which would also be wrong, but telling.
Try learning about what fascism is, that would be my suggestion. With a book, while putting your phone away.
[1] https://www.them.us/story/elon-musk-x-ban-users-cisgender-sl...
[2] https://www.advocate.com/news/x-bans-deadnaming-misgendering
pretty much exactly what i expected, you ignored all the YESs and called me dumb and made a ton of assumptions. You dont have to fit a definition 100% you're going to hang on to all the tiny little details and say wholey that he's not a fascists. I think he hit's the major point of the definition. Pedantic bootlicker. The reason I didn't pick up a book at your request is because I'm not wasting my time with you. And, no I didn't read the article, you did and linked it. I only read the definition.
Fascists do claim to support free speech while wielding the apparatus of state to ensure that any speech that they disagree with is suppressed, eg on college campuses.
That's not a rebuttal. Claiming to support is not support.
Is Musk wielding the apparatus of state to ensure that any speech that he disagrees with is suppressed, e.g. on college campuses?
> Is Musk wielding the apparatus of state to ensure that any speech that he disagrees with is suppressed
He does not need to do that, he has his army of people like you that go around on all socials and wherever you can to post praise, defend him and attack any critics as he was some sort of uber genius savior of mankind. The Musk cult is real.
So... are you arguing that Musk is an opponent of 'free speech' because... a bunch of private citizens argue with you when you speak critically of Musk?
The whole thread (and to be fair, Musk himself most of the time) has lost the plot, since the phrase "freedom of speech" is about the government not restraining or punishing speech. Honestly, what Twitter allowed or not before or after Musk is entirely irrelevant to the concept of free speech. Twitter/X is still just some guy's website.
Musk explicitly centered his DOGE effort on rooting out “woke” programs. This included pulling funding from independent research under the auspices of various government departments.
The Trump administration spent considerable effort strong arming colleges on account of speech that they did not like. Even if you subscribe to FIRE’s “Chicago Statement”, you should find the government interfering with colleges like this a break-glass scenario. “Individual Rights in Education” surely include the right to hold beliefs and teach in a way that the current administration is uncomfortable with.
So yes, these are fascist activities wholly consistent with the other fascist activities that MAGA movement is engaged in.
> Be kind. Don't be snarky. Converse curiously; don't cross-examine. Edit out swipes.
> Please don't post insinuations about astroturfing, shilling, brigading, foreign agents, and the like. It degrades discussion and is usually mistaken. If you're worried about abuse, email hn@ycombinator.com and we'll look at the data.
https://news.ycombinator.com/newsguidelines.html#comments
I'm not sure HN is for you.
uh... twitter
I feel like it has to eventually become X. Just because he’d find it funny for the stupidest people in the world to post about how he’s turned Twitter into a $2t or whatever company.
> Does the combined market cap count? That would be wrong.
It counts and it's not wrong.
Sure, the Tesla award takes into account any M&A but growing a 2T company to 3T is a 50% increase. While growing a 1T company to 2T is a 100% increase so it's expected to be easier for him to hit his award targets with the companies merged as opposed to not merged.
If we combine the market cap of the entire S&P500 we get close to 70T. That doesn't mean any of those individual companies are any more valuable or any of the investors are any richer. It makes no sense that Tesla shareholders would be ok with paying out a performance bonus just for M&A that doesn't grow the value of Tesla and would just dilute their shares.
Well yeah it doesn’t make a ton of sense but the shareholders already cast their vote on this. I’m not sure what they expect would happen but I also don’t think they care. Elons shareholders are a special kind of financial sycophant.
Most are tied to operational milestones as well if I understand correctly
As pointed out by somebody else [1]; Operational Milestones are waived if SpaceX buys Telsa.
[1]: https://news.ycombinator.com/user?id=antasvara
Thats not what it says. My understanding is SpaceX milestones stay intact.
For Tesla ops milestones they would be waived, this is fine because accusition is a chance for Tesla shareholders to cash out at that price.
We're talking about his Tesla grant so SpaceX's grant is irrelevant.
The text written in the link you did not read, states "Notwithstanding Sections I, II and III above, in the event of a Change in Control, the Operational Milestones shall be disregarded".
The link supports wxactly what I wrote. Not sure where your confusion is.
I understand Tesla acquiring other companies counting to the cap, but Tesla _being acquired_ by another company, why would that count?
Because that's the grand plan from the start. You don't put a sentence like "In other words, upon a change in control where Tesla is acquired, vesting of milestones under the CEO Performance Award does not require the achievement of a matching Operational Milestone." if you didn't think Tesla could be acquired.
The goal is to give Elon his 2012 award package but making it his 2018 one.
It the shareholders' value that counts.
And that doesn't get any more by 50% increase with the same dilution.
the problem is Tesla's board is controlled by musk
No. He is making it an AI company. The prospectus makes that clear. Everything is in service of training and deploying AI. Twitter is data, distribution and marketing, space x is distribution with data centers and internet in space. Cursor is training data and hostile distillation.
That prospectus didn't make anything clear. It was pictures of rockets and rubbish about "the light of consciousness". The only real information was buried deep in the middle and it showed a company with poor finances and no clear path to success. Certainly not something worth the likes of Amazon.
I agree with you. But also think the strategic positioning of being fundamentally AI was clear.
Unfortunately computers generate heat, so while Elon can build a business running on hot air on earth, he will not be able to do so in space.
Are we to assume that nobody at the company who had launched more satellites than the rest of the planet combined is aware of this problem?
This is very far from an obstacle, space-functional radiators and thermal management systems do exist.
How much mass do we need to put in orbit to radiate 10 GW of data center heat? How much rocket fuel does it take to lift the rocket fuel required to lift that mass? Which part of this is cheaper or more effective than building data centers 100 feet below the ice in Antarctica ? Other than serving workloads that are already in orbit, I don't see the point. How much addressable market do we have for AI compute that is already in orbit?
The appeal is solar power efficiency gains. And yes it is a notable amount of extra mass, but not a deal-breaking amount. Latency is pretty minimal regardless of location (sub-100ms) which is not nothing but it's minor for things like serving AI inference.
Not saying I think it's the best idea, but it is theoretically feasible.
When you say it is not deal-breaking - have you looked at any math? Real world data finds we need about 6kg of radiator surface to radiate 1kW. [0] . A 100MW datacenter would need about 600,000 KG just of radiator mass. ISS was only 450,000KG and it took 13 years to assemble. Meanwhile, we're projected to launch 600X that data center capacity on earth in just the next three years.
To say that this is pure folly is an understatement. It is not folly, and it is not foolishness. It is fraud.
One of the most cited papers: https://starcloudinc.github.io/wp.pdf
Does not contain the word "nuclear". Instead its just breathless gushing about how much more efficient the solar power is.
[0] https://arc.aiaa.org/doi/epdf/10.2514/1.A35030
I’m not even worried about the weight and heat. How do you do a disk swap or other service on a machine in a datacenter in orbit?
KW scale radiators for GW scale compute demand. It is not based in reality.
it actually is not that impossible, and in fact can be done in a form factor not that different from Starlink satellites: https://www.youtube.com/watch?v=FlQYU3m1e80
I didn't say impossible, I said not based in reality. Why would you do this? What economic advantage is gained?
The radiators on the ISS radiate maybe 200kW all up. That's _one rack_ in a normal earthbound DC.
Under rated comment. May be legendary some day.
Radiative heat transfer exists in space. In fact, it works quite well for cooling as long as you’re in the shade.
It exists, but "works quite well" is only for a small amount of heat. It takes a big radiator to get rid of heat to ax vacuum. Much easier and cheaper to just keep the computers on earth and cool them with standard air or liquid cooling setups.
The required radiator for cooling isn’t that much larger than the required solar panel for powering the thing in the first place, and you don’t see everyone saying those are impossible. Is it easier to keep them on the ground? Obviously, but that wasn’t the claim. It just isn’t nearly as hard to cool things in space as a lot of people seem to think.
NewBird AI is an AI company which had an extremely irrational stock price bump because it's all about AI. It AI'd its AI with some more AI to AI harder with AI... and that equaled money.
If you're writing a prospectus right now and want a lot of large institutional investment you put AI between every letter. We are in a bubble, I don't think there's any disagreement about that, when the bubble will pop nobody knows - but while we're in this bubble everything is AI. I think it's unwise to read that prospectus in good faith given all the other factors (the court case against OpenAI, the race to IPO first, Google's additional stock grant, Elon's history of corporate bailouts) that are pretty plain to see.
> Cursor is training data and hostile distillation.
What is "hostile distillation" please?
So what are the rocket ships for?
To launch the data center into space and the satellites that deliver the data to users over star link. The same reason standard oil owned railways. Controlling distribution is a competitive moat no one else can replicate. Only antitrust stopped the standard oil monopoly. Not market forces.
Sounds like a convoluted and expensive way to pollute our skies for generations to come.
maybe he secretly wants the kessler effect?
Counterpoint: what if he lied in the prospectus? With Elons proven track record of misleading investors and customers I think that possibility has to be considered.
I agree with this, but it seems so crazy to me. How can money be a motivator when you're that rich. I'm not even "rich" but I'm already at a point where money is far from my #1 motivator.
I LOVE puppies, but if I had a trillion of them the last thing I'd want is another puppy.
An interview I recently read with Seth Rogen was very illuminating (from the NY Times):
"You know how every once in a while you read one sentence and it snaps your whole perspective into place? I remember reading that book “Going Clear,” about Scientology, and there was one sentence about how if famous people aren’t treated in a certain way, it makes them think they’re not as talented as they wish they were. Like, if I go to a restaurant and I have to wait 20 minutes for a table instead of them just seating me right away, am I not as talented as I thought I was? If someone has a nicer hotel room than me on the press tour, does that mean I’m not as good an actor as I thought I was? I think that’s how a lot of famous people interpret how they’re treated."
I think the same applies to Musk. The money is a proxy for how much everybody thinks he is a special genius. Anything in his life that makes him feel less special requires more validation that he is, and money is the easiest validation he is able to acquire.
It's a game to him.
He won.
Let's move on.
Did he really win? His kids all hate him, his wives all left him, he has no real friends. Anyone he interacts with is purely after money and power they can get from him. He spends all day being angry on twitter.com.
Is that winning?
By his metrics? Probably.
I think the point of the Rogen quote is that there's no real way to win. This is not cope by the not mega-rich and famous. It's an inevitable result of wanting universal acclaim.
Rogen quote? I don't know what that is.
I think you're over-complicating this.
It's literally just Score Points.
That's all he cares about.
Uh, click the root button on your comment.
Oh! haha, sorry, it was late and I was a couple of drinks in when I first responded.
D'oh.
Musk has more money than most of us would dream of, but the game isn't over until it's over.
Speaking just for myself, I've lost respect from Elon Musk. I admire Musk's accomplishments, especially Starship and the Falcon rockets. But I don't respect Musk's personal judgement, his moral integrity or his ethics.
He doesn't know me, and he doesn't care about my opinion (or care about ethics for that matter). But there are a lot of people like me who used to respect him and no longer do. He's surrounded himself with fawning sycophants. At some level he's got to know this, and that the people pretending to pay him respect aren't themselves worthy of respect.
Most people don’t dream of having obscene amounts of money.
Status is transformative, and addictive.
Money is power. Power corrupts.
Also, dude was raised by terrible rich people and turned out to be...a terrible rich person. Color me shocked!
Musk is probably a clinical narcissist (NPD). If that's the case, no amount of power, status, or riches (or ketamine) will ever be enough
Their motivations are often cartoonishly superficial and... well... stupid. Stupid in ways that are baffling to most people (even most other neurodivergent). The kind of stupid that drives somebody to secretly pay pro-gamers to play games for them so they can pretend to be a pro-level gamer, only to then expose their own fraud by playing the game themselves on a live stream, without knowing how to actually play it. And then pretending to have connection issues when people start noticing.
I have no trouble believing Musk has simply internalized the identity of being the world's richest man and now has a pathological need to maintain that status, no matter what
It can't and it's just not. People use the word "money" for different things. He's not doing it for another bill or a number on some screens- neither are most employees of those companies. That's just projecting values on someone else.
The things they're trying to accomplish require extreme amounts of capital.
> The things they're trying to accomplish require extreme amounts of capital.
This could really cut either way. Like assume you mean he’s trying to do something good or noble (lmao) but the other obvious way to read this would be that he’s interested in becoming wealthy/powerful enough to bend the will of a nation before him and burn the world for a laugh.
It’s just funny how the markets seem to want to value him as some sort of AI visionary when his companies are not even in the top 5 for AI, despite his endless resources.
The way I see it, he's already well past the "wealthy/powerful enough to bend the will of a nation before him and burn the world for a laugh" but still needs to keep working to do his (and companies) stated "good or noble" goals. Why do the fervid anti-Elon people not even bring up his many flaws or mistakes and always concoct weird reality-orthogonal situations?
Ok sure I can be more more direct: how does aligning himself to the republican party and taking a detour to unwind as many social programs as possible in any way help his stated goals, corporate or otherwise?
The only answer is the dude is a self-service piece of shit. He has no principles and is not trying to accomplish anything other that exploit for the sake of exploitation.
That's not the only answer, it's just the one you want. There was a clear coup of the Democratic Party (or just shifting of values) that was jeopardizing both the actual climate change issues and the mission to mars. You can keep trying to find ways to justify the idea that he's just completely selfish, just wants money, or whatever, but you're going to have to do a lot more work.
Assuming best intentions, making a colony on mars is gonna require money
Not even Elon is delusional enough to truly believe a mars colony is a remote possibility.
I don't think you can describe his beliefs using booleans like that: you have to use a numeric scale. It would be be correct to say: Elon would need a hell of a lot of ketamine to believe a colony on mars is a possibility.
It's not exactly like Cookie Clicker. Many people definitely like seeing the number going up, but for most people that get to that point of wealth, the goal is power that the money represents. A human being may struggle to relate to them, but they really are motivated by the sole desire to own and control everything.
You cannot get that rich without money being the motivator.
Yes you can. Money is a medium of exchange. Stock is a medium of exchange. But they buy the ability to effectuate your will in the world. If you have a vision of the world money, human capital, political power are not the measure of success or the goal. They are the instruments of executing a vision.
Some people aren't motivated by money, they are motivated by reputation, or pathology.
> it's pretty transparent that he's just trying to ball-of-mud together everything he can to hit that target and grab the bag.
I agree, and most people do too, yet he'll get away with it. We're in the kleptocracy phase of the fall.
His award is also tied to numbers of vehicles and robots sold.
So just increasing market cap won’t do it alone.
Even if he merged them, they still have to produce WAY more than they are now.
That’s not how it works. There’s provisions to adjust the targets if there’s M&A.
Yep, bundle it all together, rename the whole company X (for added confusion), get a few final pay outs in the hundred of billions if not trillions and then Elon flees to Argentina as the whole house of cards crumbles.
Does anyone really believe he’s doing all of this just to sell out and live on a beach somewhere? This dude sleeps at his own factories. He literally works like 24/7 and has no personal life. If this was all a cash grab he’s had dozens of opportunities to cut and run with well beyond F U money. If you wanted to scam people there are a lot easier ways to do it than repeatedly founding revolutionary technology companies.
I’m not denying that his companies are awash in zany financials, but I don’t think that’s ever been the point
The optimistic take is that he is doing all this to achieve his space goals, that the ends are justification of the means.
But when he is interested in actively interfering with elections the world over and probably spends more time on twitter than anything else, I am not so sure.
The thing with the sell out routine is that they think it is all working until it suddenly isn't. The money isn't about living the high life it is about having the high score. Bezos is probably trying to figure out how he can take the No 1 spot right now. If Elon is smart, we will cash out right before the fall.
gamblers who like to martingale bet strategies often run up huge sums without ever doing the sane thing and cashing out
Elon's not going to work 24/7 through the upcoming depression
Elon flees to Mars*
Everything will be folded in "X" eventually, anyway.
> Stock that may very well be near its zenith when this closes (or maybe not. This is so far past fundamentals it’s impossible to tell).
Nah, not in a bubble or anything at all.
"The market can remain irrational..."
I called bubble on Tesla and Bitcoin both over 10 years ago.
Still going strong.
It's a whole lotta chaos, fueling some pretty messed up stuff. It is hard to watch.
Great quote.
It's easy to predict what will happen. It's hard to predict when or the second and third-order effects.
> They’re spending Monopoly money.
They also seem to be desperate to buy their way into a monopoly, even though the company itself has a long track record of failing to deliver anything noteworthy.
FOMO; Apple grew to one of the biggest companies over time and people were like "h*ck I should've bought it years ago". Tesla then seemed somewhat similar - also 5 years ahead of the curve for EVs, even if they reinvented things that didn't need to and their construction was substandard, so one group of investors bought and boosted the stock to impossible heights. Now SpaceX is the next one, and people want to get on board.
I don't think anybody actually believes SpaceX will be worth it. I think some still believed Tesla would be, but their competition has caught up. Everybody is just there to ride the wave of mass hype and FOMO, thinking everybody else is an idiot.
This has been coming in a few waves over the years in Tesla, cryptocurrencies, the Everything Bubble, etc.
Yup - given that it’s $60bn of stock, now is an excellent time to do it, as the current valuation isn’t even irrational. And I say that as someone who believes they have great long term prospects.
> It’s all in a stock that may very well be near its zenith when this closes
No, it is not. This is not legal or financial advice, but I believe the stock could easily rise 2.5x - not because of its current or future financial condition, but because it is run by what may be the most skilled fraudster our planet has ever breed. Charles Ponzi himself couldn't have pulled this off. While most CEOs are careful about what they publicly say or "predict," Musk's companies are fueled by increasingly fantastical projections. The consequences have amounted to some $1.5 million in SEC fines that, relative to the value created, were negligible - and that was under the previous administration. The current one won't be any more aggressive. The closest comparison I can think of is Trevor Milton and his famous "electric" truck that was filmed rolling downhill under its own momentum. He went to prison for that, although he was later pardoned by Trump. Many people have lost fortunes betting against Tesla based on fundamentals. SpaceX's shareholder list includes so many influential and powerful names, including people closely connected to the current administration, that I find it hard to imagine the stock being allowed to fail in any meaningful percentage. Obviously I'm exaggerating when I say the government would send agents door-to-door to collect valuables from American households to plug any hole in the balance sheet before allowing the stock to fall significantly. But that's honestly closer to how protected I think the company is than what traditional financial analysis would suggest. I'm nobody special, just someone with about $1.8 million in a stock portfolio. Yet this thing called SpaceX stock gives ordinary investors like me a chance to ride alongside the biggest players on their way to even larger fortunes. They are guaranteed not to lose money, and to me its not personal.
I mostly agree but it still seems like a huge risk to play along with the memestock-ery. I have no idea how long this can be sustained. Like, would the upcoming elections going a certain way cause things to unravel? Or, say, the Anthropic IPO tanking and shaking investor confidence (highly doubtful, but a relevant example.) Or would some other unrelated event essentially cause a gigantic rug-pull?
And this is before getting to worrying about what kind of large-scale market manipulations we're up against (e.g. relevant example: https://www.forbes.com/sites/hershshefrin/2025/04/05/signifi...) Compounding this is the fear that even the indexes seem to be compromised (fast-tracking SpaceX etc.) I'm paranoid but at times the entire US stock market looks like a gigantic pump-and-dump. (Seriously, some of the stocks, like NET, rise and fall with highly predictable regularity.)
I can't tell, and I know I'm not nearly smart enough, attentive enough, and definitely not well-connected enough to have the perfect timing required to survive a sudden shift or whatever game it is the big players are playing, so I'm just trying to play it safe (which is also hard, because everything is impacted at this scale!)
[dead]
The total addressable market (TAM) for SpaceX is finite. There are only so many nation-states and large corporations that want to launch payloads into orbit.
And even if their internet service provider is uniquely capable for now, it only fills a strategic need for certain customers.
So instead, Musk and Co. need to find bubbling market trends that look like they will have huge gigantic TAMs to justify the potential growth of this company.
All markets are finite. But you're thinking too finitely -- remember that there was a proposal to use Starship (BFS?) as a point-to-point method of people transport too (London to Sydney in under 50 minutes I seem to remember).
You also have other services: Starlink is an obvious one they're pursuing now, but there's many other things that they could branch into with no effective competition right now, from harvesting resources such as Helium-3 to Rare Earths (ironic name), to... (thinks for several minutes) banishing people to the Phantom Zone?
But you get what I mean, it's not just about rockets, it's about the things cheap and reliable rocketry enables.
Musk is a genius creating really exciting ideas. No doubt about that.
But as they say,"the devil is in the details"
- Can Starship transport people from London to Sydney safely economically, compared to Boom, which is working on a supersonic passenger aircraft ?
-Why can the boring machine dig tunnel at much lower cost than it's competitors? Maybe it's because the everyone else tries to dig tunnels for trains, which have a much larger diameter than Musk's boring machine, which only fits his "Teslas at a tunnel" concept? And it might be a good idea. Worth a try. But be honest about it.
-Sure, data centers in space probably have some great uses, and I'm happy he's trying, but will they ever be more economical than deploying servers on the ocean? On countries with very cool climate?, powered by new energy technologies?
Boom is not a serious endeavor.
They have not committed to an engine. There is no engine commercially available to buy and no engine producer has committed to creating one.
I'm not sure I understand. From a quick google search it seems like they are designing and manufacturing it themselves (with a few partners). Am I missing something, or is your information out of date?
The industrial bases of entire nations military aviation have been taxed for years/decades to produce supersonic capable engines simply for high maintenance fighters, now do that for affordable maintenance low headcount passenger jets. I'm not going to say impossible, but it will take more than VC money and anonymous 'partners' to convince me its likely.
For it to have a chance the US gov. needs to do IP transfer and pair them with NASA for test as they did with SpaceX, more realistically simply twist Pratt&Whitneys arm a bit...
[dead]
Boom is planning to operate at Mach 1.7 (approx. twice that of current aircraft) with a range of about 4250 nautical miles (4900 miles, 7900 km). Over land they're only going to fly at Mach 1.3 to reduce sonic boom effects.
That's not enough range to do London - Sydney, it's not enough to do Los Angeles to Sydney either, or Los Angeles to Tokyo, it's basically a replacement for trans-Atlantic flights only because even US cross-continental Mach 1.3 is only about 50% faster than a 737 (3h vs. 5h)... It's pretty much geared to the prestige market only.
Under an hour to anywhere on the planet, meanwhile, is absolutely someone would be a premium for -- and they'll do it for most/all long haul flights if it was available.
-And what is the safety data ?
-Musk floats this vision since 2017. Why doesn't the company have a project to seriously develop it?
Starship isn't a proven manned transport vessel yet. And Boom was founded in 2014.
If I was to hazard a guess, the BFS concept is 2035 at the earliest, most likely 2040 -- and it'll be at a much smaller scale than previously advertised.
Do you have any information to base that 2035/2040 prediction on? is there any information to base a negative prediction on?
The phrase "if I were to hazard a guess" should tell you all you need to know. All I did beforehand was update my existing knowledge by reading the Wikipedia article at https://en.wikipedia.org/wiki/SpaceX_Starship#2027_and_beyon...
SpaceX also wants to put data centers in space. That's the big market for SpaceX and how it ties into AI.
Which hasn't yet been proven to be either technically or economically viable, even on paper. It's a pipe dream.
The cynical viewpoint is that this is Elon capitalizing on current datacenter hype to inflate SpaceX's valuation based on theoretically overcoming tremendous amounts of hard physics problems, over the next 5-10 years. As he did with FSD, Boring Company / Hyperloop, Twitter, etc.
We've been through it over and over. "Tesla is not a car company it's a X company" where X is the current trending theme.
So far, Tesla has been a blockchain, energy, robotics, and now a compute company.
Neither was reuse of rockets and I remember the ex-boss of Arianespace laughing at those bozos of SpaceX who try to pretend that they are a serious space business.
Musk made some bad bets, but also some good ones (Falcon rockets, Starlink) and some at least promising ones (Starship, Neuralink). And Twitter bought him enormous political influence - I wouldn't consider this a failure either, from the realistically-cynical point of view. That cannot be measured by revenue alone.
Twitter was a "success" for musk sure, it was also a catastrophic failure for the rest of western civilisation.
Musk has been coasting on his successes from 10+ years ago. He has nothing good to offer anyone in 2026.
I think the jury is still out on the impact of all the other social networks.
If the connection between fast falling birthrates and smartphone addiction is proven, the total global loss of life (in this case, never born life) due to the products that many of us here helped craft into perfection may rival that of Mongol expansion under Genghis Khan, and Twitter/X is hardly the worst offender in that group. Even Twitter's impact on the balance of left/right politics in the world is relatively transient and small when judged against this horrible development whose aftereffect will stalk the world for a century.
Yes, it wasn't an actual intent of people like Zuckerberg, but as far as catastrophic failures of civilizations go, they don't have to be intended.
Falling birthrates predate smartphones. Where did you come up with that lol
Look at the underlying mechanisms.
Couples deciding to have fewer kids does predate smartphones indeed.
Fewer couples even forming and more people being lonely is strongly suspected to be a smartphone aftereffect.
>90% of the launch market and a great WORLDWIDE isp.
Then there is the role of Starlink terminals in an actual war.
The losses caused to the Russians by drones using Starlink for connection are pretty painful, and when SpaceX switched off non-whitelisted terminals in the theatre of war, the Russian army was thrown into disarray due to sudden failures of communication between their units. AFAIK they haven't yet fully overcome that problem.
The Russians certainly wish to have something like Starlink right now, in 2026.
In the first weeks of the war, Russia targeted communications infrastructure in Ukraine. They performed cyberattacks against Viasat, rendering thousands of satellite modems unusable. They performed destructive cyberattacks against Ukraine government and telecommunications targets. Cell towers and fibre optic lines were targeted, as well as electricity infrastructure. And they started aggressively jamming wireless communications. Much of this was disastrously effective.
Starlink was one of the only reliable ways Ukraine had for remote military communications, without which Ukraine would have not been able to defend its territory nearly as effectively. Though it's impossible to know, it's plausible that Ukraine might not exist today but for Starlink.
> He has nothing good to offer anyone in 2026.
Falcon launches cost dramatically _less_ than comparables like Ariane. In Fact, Ariane had to beg europe of subsidies to keep the program competitive.
Meanwhile, Starship is well on it's way.
You don't know what you're talking about.
Wait, you think Elon solved these problems?
Personally, I don't think so, but let us at least try to be consistent.
"If an unpopular person's corporation C succeeded at activity X, it is the success of the regular employees and everyone but him, but if his another corporation D failed at activity Y, it is solely his responsibility and shame (if not a proof of outright fraud)" is a classical emotionally charged double standard.
Do you think there's a single Apple fanboy who thinks that Steve Jobs ever had any novel insight in programming, or had any novel insight in circuit design?
Now, with that answer in mind, allow me to contend that there isn't a single Elon fanboy who thinks that Elon is personally inventing automotive and rocketry technologies out of whole cloth.
What both men did well is identify promising unconventional technology pathways and steer capital investment towards them. Jobs had a knack for understanding computers as a consumer product, and for communicating the value of new products. Musk has a knack for understanding the limits of physical engineering, and the wealth (and appetite for risk) to spam the right "build" buttons endlessly.
Beyond a narrow range of remarkable competencies, neither are particularly interesting persons. I wouldn’t look to either of them for takes on sociology, politics, biological sciences, philosophy or chord progressions.
Like it or not he founded (sometimes in-part) but drove these companies to success by demanding deliverables others thought were crazy.
For most of his sucsessful ideas he had sophisticated investors, VC's, to judge the idea and take the bet(at an early stage).
Would any VC(one without a conflict of interest) invest now, for the long term, based on those visions?
Do they, really? Because putting data centers in space would mean multiplying the infrastructure cost by a few orders of magnitude, while being far, far away from cheap energy - photovoltaics would work, certainly, but it will take a lot of it, and it's not like you can just slap panels on the roof - easy cooling, and people.
It's a ridiculous idea, and I don't believe it's what they are really pursuing.
I still have no idea how that would work. Imagine launching an entire data center building into space, and then imagine also launching a solar array to power it, and then also launching a gigantic radiator to cool it... and the radiator is full of some kind of liquid that can never leak even though it's in a vacuum.
Like, sure, but also, that seems like a lot of work, a lot of extra cost, and a lot of risk, all just to avoid building it in Kansas.
It was also hard for many people to imagine a reusable booster, a belly flopping Starship, catching the largest booster ever built with “chopsticks”, a 10,000 satellite constellation, etc.
Orbital compute is technically very feasible. We’re not talking about a datacenter-sized structure, but a lot of rack-sized satellites connected by laser links. SpaceX has gotten pretty good at building, launching, and managing large constellations.
Economically it obviously it has challenges, but there are some advantages (6x solar production, free real estate, less regulation, arguably simpler cooling) to balance the extra costs (launch, radiators, lack of access for maintenance, limited lifespan, etc).
>It was also hard for many people to imagine a reusable booster, a belly flopping Starship, catching the largest booster ever built with “chopsticks”, a 10,000 satellite constellation, etc.
I don't actually think this would be hard to imagine. I've been a huge fan of Space X since it's launch exactly because these types of things do seem feasible because they save so much of money if they are achievable.
A moon base with a secondary launch site, yes. Mining asteroids for precious metals, definitely. I'm not some Luddite.
My only point here is that you can build a data center on the ground trivially easily. Any data center that can exist in space could much more easily exist on the ground... where you can update it and fix things that go wrong. The only issue is politics. I'm entirely happy to be wrong here. If someone can explain the thesis, I'd be happy to get on board.
> explain the thesis
It's all laid out in this insufferable a16z piece if you can see past the ego-stroking.
https://www.a16z.news/p/spacex-and-the-sentient-sun
Solar power is already one of the cheaper (and cleaner!) forms of power generation. In dawn dusk sun synchronous orbits where satellites are always fully illuminated the panels will produce around 6 times as much electricity as those on the ground. And you don't need batteries to operate 24/7 (and as a bonus, the satellites will follow the work day demand curve through the day, reducing latency, at least for some people).
A fully reusable Starship will drastically change the economics of both initial launch as well as maintenance. I expect it will become more feasible to send vehicles to refuel/repair/replace components and keep satellites flying longer. Especially for orbital compute where there will be relatively few dense orbital planes. SpaceX showed modular servers in their video https://youtu.be/k3Un1TizSNg?si=14-bjxXkiyM6cxpg&t=36
First off, let’s not pretend rocket launch dependent solar is “cleaner.” Be reasonable. Solar + saline batteries is pretty damn clean.
Yes, I watched the video. It’s talking about solar and radiators. I agree, if we can solve solar and radiators, it’s feasible tech.
I’m still not entirely sure it’ll be competitive with solar data centers on the ground. I realize I’m no expert, but it just seems like a bizarre way to do computing.
Everything I like about Space X is about doing things that you can't do on earth, because you can't do them on earth feasibly.
> Everything I like about Space X is about doing things that you can't do on earth, because you can't do them on earth feasibly.
I agree those kinds of things are more exciting, but the other angle is SpaceX needs an excuse to do lots of Starship launches, just like Starlink has done for Falcon 9. If orbital compute can at a minimum be profitable enough to pay for the launches it will help bring the cost of Starship down and the reliability up, allowing SpaceX to do more with Starship.
(Of course if orbital compute merely breaks even then maybe the current valuation isn’t justified)
> catching the largest booster ever built with “chopsticks”
I've heard this often. It's not what happens. More correct to say that engineers (not Musk) got a booster to land in a pre-specified spot. The "chopsticks" aren't waving around to catch anything flying by. The booster comes to them.
LMAO, ok, how about “precisely landing the largest flying structure ever built such that it could be grabbed by two mechanical arms attached to the launch pad”?
That will make it clear it’s not actually a giant Mechazilla robot reaching out to grab the booster using literal chopsticks where ever it happens to come down.
Perfect!
There's no laws in space, which is the key.
Completely ignoring all the ways in which space data centers are fucking infeasible for a moment.
Just because there are no laws in space, you still need to launch your rockets from a place that has laws.
You should actually stop and try to imagine it, or failing that, read some of the proposals from companies who want to do it.
None involve launching buildings.
You can look at their models for comparisons with Kansas. There is literature and papers you can read about this. Some go back decades if you include space power transmission, which are related.
That's the cool part, scoofy. You don't need to understand how it will work, you don't need to take any physics classes, round up enough investors, seek out and explain the basic ideas with the people who will make it happen, or invent anything new. Nor do you need to understand political sciences, taxation, jurisdictions, supply chains, or anything else needed to understand the question behind the Data Centers in Space solution.
You aren't even being roped into it with taxes, nor do you have to buy a single share. Other than willingly reading about it on whichever news sources you choose, your observed life will not change a single bit.
You can choose to seek out that info, or you can remain blissfully ignorant. But please don't join the online cacophony of people polluting the threads thinking everyone wants to understand just how ignorant they are.
I get it, I really do. It's a hard task and you don't understand it. But WHY do you feel the need to share that you don't understand? Do you think it makes you look smarter? Do you feel like you fit in more now? If you seek to understand, why aren't you asking questions??
>You aren't even being roped into it with taxes, nor do you have to buy a single share.
Because of the eventual index inclusions, and insane market cap, this affects nearly everyone with a retirement account. Unless you aren't tracking big indexes for some reason.
And some of our compute will run in space, the point is that it's all happening in the background. Most people have no idea how their retirement accounts work beyond knowing it's a bunch of companies pooled together. You don't need to understand how stocks get traded through Alternative Trading Systems, how the companies can decide to take profits vs paying out dividends, etc. A lot of the information is freely available, but you don't need to understand it or take any action.
It's much more likely SpaceX will continue building more ground data centers and using their sat relays to make global connection faster than ground connections can allow.
A great business for the rocket logistics company since radiators are a thing.
A less good business for the data center company.
xAI could tie in just fine without Cursor in the picture.
yes musk said that, but that's retarded, a statement made to fill as many bingo spaces as possible
Yes but the statement is in fact a milestone to meet in order to vest Class B stock options, specifically SPCX needs to put 100 terawatts of compute [1] in outer space and beam it back to somewhere, my guess is Earth.
There's even more rewards for putting a million people on Mars and reaching a market cap of 7.5T by a certain date. Oh yeah he has to stay employed too.
From the SEC Form 3 filed June 12th: 1) This Form 3 does not include 1,302,072,285 shares of restricted Class B Common Stock issued to and held of record by the Reporting Person, which may be voted by the Reporting Person, and the vesting of which is subject to the satisfaction of certain performance and other conditions. 1,000,000,000 shares of restricted Class B Common Stock vest upon (i) the Issuer's achievement of specified market capitalization milestones across 15 equal tranches ranging from $500 billion to $7.5 trillion, with each milestone reflecting $500 billion in additional valuation, and (ii) the Issuer's establishment of a permanent human colony on Mars with at least one million inhabitants, in each case, subject to the Reporting Person's continued employment ("SpaceX CEO Award"). 302,072,285 shares of restricted Class B Common Stock vest upon (i) the Issuer's achievement of specified market capitalization milestones across 12 equal tranches ranging from $1.065 trillion to $6.565 trillion, with each milestone reflecting $500 billion in additional valuation, and (ii) the Issuer's completion of non-Earth-based data centers capable of delivering 100 terawatts of compute per year, in each case, subject to the Reporting Person's continued employment ("AI CEO Award")
Why is this the case? I want him to be correct, but I am also skeptical.
> I want him to be correct
Why?
If you're curious, I explained why space data centers are such an irredeemably stupid idea in my eyes a few comments up.
Why? Why not? Why wish failure upon someone who is trying to do something “good?”
I haven’t read your other comment yet and plan to, but it seems weird to seemingly wish ill on someone’s grand plans.
The way people just casually use that word again now is so sad. And I don't even mean that in an "I'm offended" way, but more of "I'm embarrassed by the way you're trying to be offensive" way.
> I'm embarrassed by the way you're trying to be offensive
Oooor, try this one on for size:
What if they're not out to cause offense and the malice you impute is just an illusion under which you yourself are laboring alone? What if it was a well understood and not particularly offensive vernacular usage from before people decided they ought to spend their time being offended on behalf of hypothetical listeners?
Why use a word that has some offensive quality to it when other words would be just as effective in communicating whatever you're trying to communicate? You're actively making a decision that you know will cause some level of offense. So the only conclusion I can make is that some level of offense is intended.
In 2004 I used to volunteer as a tutor at an afterschool center in a low income housing project. One day a middle schooler was complaining about how much homework they had and I ribbed them a little, "oh, poor baby."
They were stung. "I'm not poor!" I felt so bad about it that it's stuck with me all these years. Does that mean because I've seen first hand how hurtful it can be that I should chide people whenever they use the P word?
"Chide" is not the word I would use because there is a very obvious difference between the offensiveness of "poor" and "retard", you obviously know that. But yes, if I heard another volunteer at a program for low income kids use "poor" in that offhanded context and I saw the pain it caused in those kids, I think it's reasonable to take that volunteer aside and say "be careful using terminology like 'poor' as it can be surprisingly hurtful to kids that are self-conscious about that". You can do that sort of thing in an informative and compassionate way without being "chiding".
And that analogy isn't even accurate because I'm not the one informing you that the word can be hurtful. You're using it already knowing that. So a better question is did you continue to say "oh, poor baby" to the kids who were hurt by your original comment?
Why would you think I'd continue to say it after realizing I'd inadvertently hurt the kid's feelings? You are making assumptions of ill will from me in the anecdote I shared just like you are making assumptions about the OP intending offense because you didn't like their word choice.
>Why would you think I'd continue to say it after realizing I'd inadvertently hurt the kid's feelings?
I don't think you would continue using it, that was the point I was making and it sounds like you now agree with me that we shouldn't be knowingly offensive.
And my point is that I went out of my way not to use it any more in that circumstance.
Yet in the years since, I still talk about poor decisionmaking, poor luck, poor performance, and poor word choice. Because it would be poor logic to go through life auditing everything I say just in case a middle schooler with a somewhat poor vocabulary might mistake my meaning.
Which brings me right back to "there is a very obvious difference between the offensiveness of 'poor' and 'retard', you obviously know that."
"Poor" has non-offensive uses so you can continue to use it in other ways. You don't need to advocate for the non-offensive uses of a slur. You know regardless of the context, some people will be offended by its use.
Mercy me, a slur? Retarded also has a non-offensive meaning. It just means slowed.
In fact, while I wasn't around at the time I'd wager that "mentally retarded" came into an official usage specifically because it was a clinical, sterile, bloodless, and utterly anodyne descriptive term. Moron, imbecile, and idiot all were once clinical terms. And then people throw them back and forth at each other to call each other stupid, they gained an offensive connotation and new terms were needed.
In 20 years will you find it absurd if people say that "differently abled" is a slur? Will you say "this is nuts, we literally came up with that term to avoid offense?" I will!
Try calling the next Black person you see “negro” or “colored”. The idea that nomenclature can’t evolve is bizarre.
I just showed this to a Black person and they said it was regarded, whatever that means.
More seriously. Is "moronic" okay? It's just an ever so slightly more archaiac way to say retarded. The meaning, the negative connotation, the level of offense—it's synonymous and analogous across the board in its time.
Is it okay because "nobody means it like that, they're just using it as a synonym for stupid?" If so, congratulations, you now understand the other side of the argument better than when you started.
Yes, I think there is an intention to cause or risk offense. On the other end, I think there is an intention to be offended and failure to mitigate.
It is a fairly common conflict that arises as a flashpoint in many areas. Different social and legal theories come up with radically different standards.
If someone has a cold, should they not go shopping out of caution for others? If someone is immune compromised, is it their responsibility to take precautions in a store?
Do you just.. never swear?
It takes a big ego to look past fundamental attribution errors.
https://en.wikipedia.org/wiki/Fundamental_attribution_error
oh, no it's exactly as jimmy valmer puts it, there is nothing against mentally disabled people, it is just it's something so stupid that one can't even decide were to start to describe all the points in which is stupid, so stupid doesn't possibly cut it
If you can't think of any other intensified synonyms for "stupid", you just might be...
Ironically, we can thank Elon Musk for that too.
edit: Gross that you're being downvoted. HN crowd needs a serious look in the mirror.
Bringing back and pinning the word to not derail the discussion of "mental illness", "mental handicap", "slow learner", etc or its use as an offensive?
I think the main issue is that no matter which word/phrase is used, some people will use it as a slur, and changing it so often causes more issues than it solves.
For what it’s worth, I’m not trying to be offensive or edgy when I say that word with friends. “The grass is green and that thing (random topic) is retarded.”
You know that many people are offended by that word and yet you use it anyway when other words would get the exact same message across without the offense. The only reasons to use that specific word are either the desire to cause offense or to revel in the possibility of causing offense.
That is your opinion, and you are entitled to hold it.
exactly.
There's no need to use slurs.
Musk himself has identified as such: “For the record, I am a fat retard.”
So, that is in fact, his word.
If someone refers to themselves by a particular slur, that does not grant you any social leniency to call them that too. Consider that exact situation with any other particular slur.
I have not called musk redarded, i have called the idea retarded.
Same as master branch or master/slave communication idiocy. People that get this hung up really have too much free time on their hand, or have too much to gain in discussing language instead of discussing the actual problem.
For example: see how many commenters here are debating words instead of debating the validity of yet another hype-inducing value-pumping statement for clueless investors and fanboys.
On local fintech radio the week it was first announced they spent 3-4 days discussing the financial implication, then just one discussing the feasibility. Guess we're lucky we got one whole day of engineering discussing things for a change
Musk using a slur about himself doesn't grant permission for others to use it. Would monegator have used the n-word if Musk had used it to describe himself? Hopefully not. And let's be honest with each other: Musk says things like this for shock value, because it's not an acceptable word.
Energy will be the biggest bottleneck to data centers on land. Is not an issue in space. Space is the perfect env for running compute.
Space is an abysmal environment for running compute. It offers no real advantages over doing the same thing on Earth, and it's more expensive, too! Energy is far cheaper and more abundant here than in space. And get ready to figure out things like:
- Heat dissipation
- Radiation shielding
- Either the most complex in-space construction ever undertaken, or the most complex distributed computing problem ever undertaken (no, Starlink satellites aren't good enough, we're orders of magnitude away from replicating the speed and reliability of connections within a single room)
- Zero flexibility, zero repairability, zero upgradability. Either it's working, or you make it burn up in the atmosphere with no in-between. Add on that the rationality of sending mountains of precision-manufactured tech containing many uncommon metals only for them to be completely lost. This makes the pricing even worse, in addition to
- Already high costs for designing, building and launching all that in addition to all the extra weight overhead you're taking in components that don't do computation, when the alternative is building a glorified warehouse in the middle of nowhere.
It just doesn't make any sense. It's a project tied up in hype and created solely so spaceflight can be hastily duct taped to the AI investment hysteria. Ask yourself why no one brought this up before or outside the context of AI, despite the lowering of space launch prices and data centers both existing before any of it.
> why no one brought this up before or outside the context of AI
AI compute is different (slightly higher latency is fine for inference, and there's no issue for training), and there has never been so much backlash against data centers or other infrastructure buildout. If an increasingly-non-minority of politicians get their way it will in fact be cheaper and faster to get some servers shipped to space than it will be to get the permits and build it out here on earth.
Also, most Datacenter maintenance is just dealing with the problems you see with space- power and cooling. Solar is 5x better in space and a lot more consistent. Here on land they're shutting down nuclear and imposing so many new regulations on gas and coal (and now on solar and wind) that there aren't many grids that can support growth at the scale that's being requested.
Nobody is claiming that all compute is going to space, which is what you seem to think you're arguing against. There's high-dollar demand for it right now, so if we want to be multi-planetary it's the perfect time to start tackling the "compute in space" problem that needs to be solved. Or you need to prove to future people that "All compute can be forever located on Earth" which seems a lot harder sell.
> never been so much backlash against data centers
Backlash against the proliferation of space junk still has plenty of room to grow.
But it'll have to start with Sinophobia so nobody wants to touch it. LEO objects are just such a non-issue that you'd have to really be grasping at straws (though the "Data centers will consumes earth's water" people can probably fall for it)
Assume reusable spaceflight eventually brings launch cost close to the cost of fuel. This is close to happening.
The overhead of building out grid and power infrastructure on land would then exceed the installation speed and cost relative to space based deployments.
Also assume the compute that does make it to space has a short shelf life anyways so lack of ability to repair is a non issue. As we scale manufacturing on land this will increasingly be the case.
China has already run experiments and served models from space, so we know the heat dissipation equation is solvable.
Finally you’d arrive at a similar model that’s already proven successful with Starlink but applied to serving inference.
The key question is speed to scale new deployments to meet demand. If the markets demand is near infinite, they will choose to fund space based deployments over slower land deployments.
That makes no sense to me. You've cited one overhead for building on land but just didn't bring up the billions of overheads for deploying satellites. The associated costs of launching this stuff into space blows the price of boring old power grid infrastructure out of the water, even with launch prices assumed to be the cheapest they can be. For instance, the extreme additional costs of equipping each payload with their individual systems of power generation, flight computers, maneuvering systems, structural components, communications infrastructure, heat dissipation hardware, radiation shielding and so much more. You have to pay the price in weight and money for all of these inefficient small-scale components for each unit launch, compared to plugging in an additional server rack and taking advantage of the economies of scale in a centralized data center that already solve all of these things. And that's before you get into the costs of designing, mass-producing and operating these things, compared to the costs of running a normal data center. And that thing about reusability that I mentioned also cuts into the margins. Data centers can keep working even if their hardware is a generation out of date, they can sell off old equipment, they can repair individual components instead of throwing out a whole rack. This is literally setting your GPUs on fire. It's just one more thing that makes it even more expensive.
Heat dissipation in space is possible, of course, but every kilogram you spend on heat is a kilogram you could've spent on something else. When you're talking about boxes that generate so much heat, you're going to need to spend a lot on that ancillary hardware in each unit that, again, makes it even less rational.
Then the concerns about megastructures or distributed computing go unanswered - to my knowledge, we simply don't have the technology for either of these right now. Starlink isn't close to solving it - the bandwidth of a Starlink satellite is nothing in comparison to the bandwidth of a single current-gen server GPU connection.
Energy is not cheaper on earth. Solar in space gets 24/7 power at a 30%+ rate vs the surface. Radiative cooling is passive and cheaper because there are no HVACs or chillers and high temperature chips reduce the need. The ISS does this already. Radiation really isn't a big issue with ECC and redundancy and optical links are fine for batch training.
The only hard part making the math work is the launch cost. They need reusable and reliable starship economics. If they hit that goal, it will become cheaper for pre-training, which is 70%+ of the budget for the frontier models.
> It offers no real advantages over doing the same thing on Earth
Abundant solar energy, free real estate, less regulation, less backlash from NIMBYs, simpler (yes) cooling.
> Energy is far cheaper and more abundant here than in space.
Huh? The sun is obviously the most abundant source of energy in the solar system.
Satellites in a dusk dawn sun synchronous orbit can be fully illuminated 24/7, so they receive ~6x more solar energy than panels on earth. They also don’t need batteries to operate 24/7, and the panels don’t need glass to deflect hail.
> Heat dissipation
Yes, it will require large radiators. They’re mechanically simpler than terrestrial cooling though.
> Radiation shielding
It turns out generative AI is somewhat uniquely robust to occasional bit flips.
- Either the most complex in-space construction ever undertaken, or the most complex distributed computing problem ever undertaken (no, Starlink satellites aren't good enough, we're orders of magnitude away from replicating the speed and reliability of connections within a single room)
It won’t be a single structure, and will only be used for inference, so latency between satellites/racks doesn’t matter.
- Zero flexibility, zero repairability, zero upgradability. Either it's working, or you make it burn up in the atmosphere with no in-between. Add on that the rationality of sending mountains of precision-manufactured tech containing many uncommon metals only for them to be completely lost. This makes the pricing even worse, in addition to - Already high costs for designing, building and launching all that in addition to all the extra weight overhead you're taking in components that don't do computation, when the alternative is building a glorified warehouse in the middle of nowhere.
I think people vastly underestimate how much a fully reusable Starship will change the economics of space operations.
Not only the initial launch costs, but things like refueling and repairing satellites becomes more economical. I wouldn’t be surprise if SpaceX sends Starships to refuel/maintain satellites to keep them in orbit longer. In fact, SpaceX’s own animation shows modular servers sliding in and out of the satellites.
Space-grade photovoltaics are >10x more expensive than ground based panels. Add some (Tesla) utility scale batteries and it can run 24/7. No need for expensive radiators or rocket launches. And personnel can upgrade the hardware every time there's a new generation of GPU's.
Putting datacenters in deserts around the equator is a much better idea than in Space. If you're really optimizing for cost that is. If you're optimizing for SpaceX meme-stock valuation the former wins
Perfect, minus that pesky law of thermodynamics.
What about the heat?
[dead]
A global centralized internet provider? I mean, just that, might be a trillion dollar company. Let alone build out datacenters? Those are not easy or cheap here on earth. You can build out datacenters at scale with minimal need for power or cooling. I think the current rate for a data center (not gpu) is 120million. Datacenters are super hip now too.
More like a giant military radar larping as an space based ISP.
https://www.youtube.com/watch?v=jbp3kdJZ1_A
Elon has always been a Department of Defense/intelligence asset larping as a capitalists. You can't exactly go to congress and get a quarter of the government budget to do globe spanning surveillance from space, or convince them to let you fund, build and maintain a platform for psyopping your own civilian population (x) So you go get your illegal projects funded through capital markets with through an asset like Elon. As soon as the department of defense started focusing on subterranean warfare and how its the future of wars Elon starts a tunneling company.
Elon is just Howard Hughes v2, they're running the same script again.
Space Opportunities + Robots (Tesla merger when it happens) + Software factory (Cursor).
It is a umbrella enterprise.
Is the TAM for airlines finite? They ship an indefinitely growing volume of cargo and people. Reusable rockets will be no different. Cargo into space, people in point-to-point orbital flights and to Mars.
Not a sound argument because you would have said the same thing before they did Starlink. "The TAM for launch services is finite. There are only so many countries/companies that want to launch satellites."
Musk would argue infinite. They literally want to create offworld colonies, with everything that entails. Obviously it's crazy, but it beats the pants off more adtech.
I'm bullish on DC in space with laser links. The whole sentient sun/railgun on the moon... hey, go big or go home. I would have probably just asked MBS for money on that one, and renamed the railgun "the line (of ketamine)".
It’s insane enough to consider undertaking the risk of a moon or mars colony. But doing so with him as the sole load bearing link in your supply chain? What if mars turns out to be woke and you’re fed into a wood chipper like USAID?
Last week a 13 year old video of ceo of Ariane Airspace got popular on twitter. When asked about spacex and reusable rockets he said: "there are only 25 satellites launched a year, every year, and that’s not going to change"
Currently a single Starlink launch is 25 satellites. And there are 100 such launches a year.
If only the shareholders had any kind of voice - if only it was illegal to issue a fake IPO where you sell an overwhelming number of shares stripped of their voting power - if only the market responded rationally to this boondoggle.
If we're going to right the ship in turn of common sense a bunch of people need to lose a bunch of money, I just hope it doesn't mostly hit passive investors and instead lands mostly on Elon-stans.
If only you there were companies who could choose which type of shares they'd like to offer...and if only people could buy the types they want from the company that aligns with their ideas...
Passive "investors" can go and invest in ETF or whatever else that does not include company shares without voting rights.
How is it legal to have different share classes? You could make 100 shares that can vote and then sell 99.9% of the company while maintaining full control. Seems strongly against the spirit of a publicly traded company
Most large companies do that these days: GOOG, META all have different share classes. Even the small startup I worked for had that.
I mean that just begs the question how we got to this point. And how far detached is the public market from real value. Many shares don’t confer actual value. Hell a lot of these companies don’t pay dividends in addition to very little or no voting rights.
Dividends are an interesting topic. There are sort of two ways to view the stock market - one is as a loan with a very drawn out repayment schedule that may be amicably adjusted for the health of the company (that one relies on two repayment strategies, dividends, buybacks and acquisitions). Once a company has issued stock the only two real reasons to care about the value is for the utility as a valuation tool which has steeply fallen off with indepth audits becoming the norm and, of course, further stock issuance - a healthy share price allows less dilution for further issuance. The other view of the stock market is just pure gambling - if you take away the disbursements and shareholder voting control then all that's left is the valuation which we can do better in different ways (and stock price is often disregarded when considering loans and M&A except when the difference can be arbitraged).
Really, it's kind of all meaningless BS, but as long as most people believe it isn't then it's a great way to grow assets by doing nothing which is an activity that we apparently want to encourage?
I wouldn’t see it as just pure gambling. I’m not a dividend queen that really cares all That much about it. I do think we are sliding along the scale from shares having intrinsic value as an ownership stake in an income producing asset to collector cards that we track societies valuation of a company with. And at some point in the next 50 years shit will hit the fan. Unfortunately sitting it out isn’t really an option either
There's some companies (Snap being one of them) where certain classes of shares have NO voting rights. If you think SpaceX is lopsided, look at Snap's shares. They have A, B, and C class. The founders of Snap own 95% of the voting through their Class C shares... literally only pre-IPO folks have any voting rights.
Musk didn’t invent the idea of using multiple share classes to ensure the founder(s) retain control of the company, see Rupert Murdoch, Google, Facebook, etc
From the regulators’ perspective: it is a risk, but you disclosed that risk in the prospectus that buyers are assumed to have read (what percentage ever actually do?), hence it is fine
Well, when you buy into an IPO, they make you sign to say you read it. So either you did, or you made a false statement on a legal document
Who has the gold makes the rules
I suppose they kind of do, they could sell the stock and drive the price down. That wouldn't force Musk to change direction, but it would hurt his wealth.
If he was smart it'd be to a very limited extent.
If you have this much wealth in something as unpredictable as the stock market I think the very wise move is to start converting it into stock backed loans[1] that you leverage to buy real assets - buy islands and small countries, buy other successful large companies and just let them keep trending slowly upwards, pick up that New Zealand bunker... just try and get your money diversified and into things that would survive a stock price collapse. Even if Elon did this with 1% or 10% of his wealth and kept the vast majority of his assets invested in this one clump of companies and they went to zero he'd still be sitting on an insane pile of cash.
1. Lets all pour out a drink to the bankers who authorize such insane loan agreements.
With its IPO, SpaceX secured its role as the vehicle for consolidating Musk's vanity businesses into one closely held public organization that can more easily convert publicity into investment and internally reallocate funds and debts based on his personal whims.
So yes, SpaceX is pivoting, but it's to no one's surprise.
So buy $TSLA is your recommendation
Buying Tesla is just buying SpaceX. When they roll up Tesla, it will be in a stock deal not real money. None of this is real money and I would argue that Musk is not the world's first trillionaire because there is no reality in which he could get that money out of SpaceX
I mean, if he wanted to sell tomorrow, who COULD spend $2-3 Trillion to buy it, and who WOULD? Anyone with that kind of money to spend today knows what a scam it is
Sure, it's reasonably likely they merge at some point but the ratio between the prices at that time will be different than the current ratio, so it still matters which share you buy today.
Especially since Musk is heavily incentivised to merge them at a price favourable to spacex since he has a much larger share
Musk may not have a few billions liquid on his cashing account, but he have something better: collateral to raise a few billions for everything he may want.
I mean as far as I know no human being is even a billionaire anyway if you only count cash. It's one of the things the "eat the rich" crowd is particularly bad at internalizing, that there's a difference between value and money that can be spent on food or hospitals.
It is much more feasible for Jeff Bezos to sell a billion dollars worth of Amazon tomorrow or Bill Gates to sell a billion dollars worth of MSFT tomorrow than it is for Elon to sell a trillion dollars of SPCX even over a year's time
I get that net worth is more than just cash, and that is not what I meant and it's pretty obvious that isn't what I meant. It may not just be cash on hand but if an asset is completely illiquid at it's purported value, is it actually worth that?
It is, because of the time component of money. Money is a way of storing the value of labor in a manner less affected by time. If you sell it all at once, its value is greatly reduced. All investments include that time component. Point-in-time net worth doesn't have much value as a measurement in part because of this.
2,000 gold bars were stopped out of iraq. and that is just what was stopped / reported.
Dictators and autocrats may or nay not have cash sitting in a bank account, but there are most likely multiple with $1B in gold.
Gold is value, you cannot take it to a grocer and buy food.
Somewhat doubtful. (The first part of your statement, anyway. Of course the difference between abstract "value" and hard, spendable "money" is a thing.)
Like Mr. Hanson said in my sibling comment, some rulers are (or were!) bound to have amassed incredible amounts of resources. For historical/non-present-day examples, consider looking into figures like Jakob Fugger or Mansa Musa.
I have no idea why the adjusted net worth of people who died centuries ago would have anything to do with money in the modern day. And at a glance, neither of your examples had billions of dollars in cash, which is the point I was making. They may have had a lot of value tied up in illiquid investments, which is exactly like Musk's valuation.
Cursor's edit data is invaluable to anyone who wants to train a coding model. Probably the best data available outside Anthropic and OpenAI. Coding models are seen by the leaders in AI as both the biggest current revenue opportunity and the best way to accelerate the progress of AI and bring about recursive self-improvement that will create superintelligence. So yeah, it's easy to see how SpaceX could see it as in their interest to purchase Cursor with 2% of their equity.
Mars was never going to happen without revenue. Starlink is providing revenue but probably not enough to build a whole city on Mars within our lifetimes. SpaceX needs more and AI is the only near-term way.
> Cursor's edit data is invaluable to anyone who wants to train a coding model.
Ok. So what prevents a company from offering a Claude Code/ Cursor equivalent, with 100% subsidised Claude (= 100% free), capturing the exact same data that Cursor does? If the data is worth in the tens of billions, the cost of subsidising the usage is negligible.
Cursor also has a large customer base including most of the Fortune 500, talent, and their own coding model and training infrastructure using their data. You wouldn't get those automatically by subsidizing Claude, and the many months or even years it would take to ramp up your own IDE/coding hardness from scratch and acquire customers to match Cursor would put you way behind in a field where the SOTA climbs every month. It's also not clear that Anthropic would let you undercut everyone else using their API; they could cut you off at any time. That would be a very precarious position.
This Cold Fusion vid covers the "pivot" nicely.
https://www.youtube.com/watch?v=FPIGu0anfAE
The video explains that it is spelled out in the prospectus that SpaceX is counting 70%-80% of their total addressable market to be AI related and only about 7%-8% to be space-related.
I saw that too, and it's so depressing. SpaceX was pushing the envelope of "interplanetary" travel/species, and to see it being reduced to a 7-8% "side-quest" :-(
> was pushing the envelope of "interplanetary" travel/species
I don't think they were. All the Mars stuff is just dressed up version of The Boring Company - a distraction by Elon to better position his other interests.
There's no money in going to Mars and there's no reason to, from a financial perspective, and Elon doesn't care about anything beyond wealth and power.
> SpaceX was pushing the envelope of "interplanetary" travel/species
We already have a perfectly usable planet. It just need to be taken care of.
Why not N+1 planets? Too many extinction level possibilities outside of our control.
Calling it an IDE is under-representing cursor
They have in-house models, and the data to train even more powerful ones. The cursor team is a proper AI lab.
> Calling it an IDE is under-representing cursor
On the contrary, it's over selling it: it's a not even a stand-alone IDE (like Zed, for instance) it's a mere fork of VSCode.
In the same sense that chrome is just a safari fork I suppose
Safari isn't even open source, and Chrome has never been a fork of Safari.
But yes Blink definitely started as a Webkit fork, and everyone would have found that laughable if someone bought that a proprietary fork of Webkit for $60B.
Isn't their in-house model just Kimi?
See here https://cursor.com/blog/composer-2-5
85% of the compute for the final model is from them, and not the base Kimi model.
That just means it cost a lot.
Does it perform meaningfully better than the Kimi model given all that extra compute? And proportionally to the amount spent?
Cursor's Composer 2.5 is one of the few models out there focusing on coding, which is the one thing most of us here want. It's pretty good! It's not near frontier level insight generating genius, but it's regarded as very capable and trustable, and is indeed a lot better than previous Kimi. It'll be interesting to compare it versus Kimi 2.7 Code, which just dropped, which is also notably a coding specifical model. I'm expecting we'll see more of this over time and I think it has huge rewards, and Composer 2.5 is early proof.
I'm not super concerned about the spend to train the model, especially given that Kimi was famously incredibly cheaply made, and given what they are competing with. I don't think that's a meaningful concern.
Reciprocally, and in far more important relevant in my humble opinion: in terms of cost to run models: Composer 2.5 is easily one of the cheapest models out there. It's fantastically cheap. It's token efficiency is through the roof astronomical. I think this training for a coding specific model has yielded something incredibly special here, and I hope SpaceXLAIC isn't the only company doing this.
That's something for us and benchmarks to decide
However it definitly isn't _just_ Kimi. The weight will be different after that 85% of extra training on top of the base model.
If those different weights are better are worse doesn't change that it's in most meaningful ways not the same as the base one.
I would encourage you to lookup their blog posts about their post training process if you want a bit more faith that they aren't running an extra 85% of compute and burning money with no-ops.
"Just Kimi" is hyperbole, to be clear.
I don't think it's all no-ops. Still don't think it's a particularly relevant model/company/product.
I'll defer the reading until I see signal that they have something worthwhile. I've watched a couple interviews and used the product, neither of which impressed me.
You don't think that a $60b valuation is having something worthwhile?
(Only half-joking…)
They use Kimi and post-train it on the same stuff that anyone with a Github dump can feed it. They aren't doing anything that you can't do yourself.
Dumping github into a model is not post training, thats pre training. And every base model already has all of github.
Composer post training is clearly very good, only second to Anthropic and OpenAI.
It does irk me a bit that they try to hide the fact that it's based on a chinese pretrained model though.
why comment on something you clearly don't know anything about? it's on-policy RL trained not just on coding text
listen and learn :)
Meh. On an outcomes analysis, I've found Cursor's delivery to be exceptionally weak.
Good luck to the alt-economy of SpaceTesla though, may all our 401ks survive.
Their “in house models” are reportedly basically just Kimi.
Replied on the other comment about this, but putting it here:
> See here https://cursor.com/blog/composer-2-5
> 85% of the compute for the final model is from them, and not the base Kimi model.
Of course they could be lying, but it seems feasible that they are adding a lot on top of this
>One of the things that makes @SpaceX so valuable is how valuable it is. The Cursor acquisition costs materially less in dilution because of SpaceX’s high valuation.
https://x.com/BillAckman/status/2066866144154161555?s=20
They are buying it with overvalued stocks, so it isn't real money. Probably the Cursor team will be able to sell it when the SpaceX stocks will be already crashed.
Why are you comparing it to building expensive modern hospitals? Why don't you compare every other tech acquisition to that? because that's not a relevant comparison, and building expensive modern hospitals has nothing to do with the goal of for-profit corporations.
SpaceX has 3 major businesses: Space, Starlink, and AI.
This acquisition helps with the 3rd one.
This is the same straw man argument of "country Z has N homeless/unhoused people, why are they building a space/military/education widget for $Y price, when they haven't found homes for N unhoused people yet?"
Except businesses tend be focused on core competencies and not sprawling in a lot of unrelated directions.
yes, whataboutism at its finest
They compared the acquisition to hospitals because "think of the children!"
Evidently, at some point money just stops being real. Also, it's not even $60b. It's a 100% stock deal.
The markets have quite literally given Elon a carte blanche. Can it last forever? I doubt it.
> all-stock deal
It's not like they could easily cash out all of those $60B. I always find it troublesome that we generally conflate cash with stonks, market caps, and such.
> It's not like they could easily cash out all of those $60B.
They probably have a vesting period of some sort (as they would with cash as well) but beyond that they will definitely be able to cash out all of their money as soon as they are allowed to.
$60B is 3% of SpaceX at today's valuation, Musk had no issue selling this amount of Tesla shares to buy Twitter. The idea that stocks are somehow not liquid is an nonsensical urban legend.
I wish we’d stopped calling him Elon like he’s a family uncle or something. I chuckled at Heilon Musk, but I’d settle for just Musk.
Back-of-the-napkin math tells me Google’s top 20 acquisitions combined are cheaper than Cursor.
Elon is consolidating all of his property into one single megacorporation because he is confident that nobody will ever challenge this, given the current political direction of the United States.
> the current political direction of the United States
These kinds of comments reek echo-chamber parroting and zero substantive research. As someone that very much enjoys and carefully follows politics, the current political direction points squarely to Republicans getting absolutely pummelled in the midterms, effectively turning Trump's administration into a 2-year lame duck. What are you even talking about?
Will a midterm pummeling change the regulatory departments that oversee mergers and anti-trust?
Obviously you're trying to be snarky, but I hope you realize that Congress does, in fact, have (fairly broad) statutory authority over executive agencies[1].
[1] https://www.congress.gov/crs-product/R45442
Can it change? Yes. That's not the same question as will it change. And that is also not the same question as "will the change result in a different posture towards antitrust."
When was the last time substantive antitrust action was taken that forcefully restructured a large company to a significant degree?
And I hope you realize that while Congress has authority over a lot of things, that authority is being routinely overridden by the current presidential administration, including fundamental things like spending and declaring war.
So it comes across as a bit foolish to assume that any Congressional authority actually exists, or will continue to exist into the future, since we have many examples now of where that authority seemingly doesn't matter anymore.
Especially since the majority of Congress is in the same party as the current President, and is making no effort not to cede congressional authority to the executive branch.
Congress isn't being overridden at all. Congress has the power and authority to rein in Trump whenever they want. They are choosing not to, because congress is controlled by the Republican party, and the Republican party is currently ducktaped to Donald Trump himself, and even barely delaying anything Trump wants gets you primaried.
But mostly they are fine with what's happening so they have no desire to stop him.
One of the things I find really interesting as a Brit is hard to put into words:
Americans tell us that we don't have a constitution — when we do, it is just not wholly written down. (It is in part).
We have a constitution that is flexible and precedent-based, but pretty stable, and it has emerged on top of the bits that are written down, and has amended them over time (for example, it is built in part on Magna Carta, but only two or three of its principles remain in law.) Notably a bit more of it got written when we agreed to be bound by the ECHR, but that was mostly absorbed into our understanding.
It has taken us hundreds of years to get to this stability, and it is defended from attack from pretty much all sides; every government risks changing it and there is pushback each time, because you can't govern if there aren't rules. The rules are precedent and convention, and there are various authorities and archives that are consulted to work out what they are if people think they are at risk.
We are regularly told by Americans that this is an intolerable thing; we need a written constitution or we can't know what our rights are!
But those same Americans, right now, are engaged in exactly this process. You have a set of written rules that give Congress power over things, and you are currently evolving a set of precedents that suggest that the executive can simply wander past them and Congress somehow shows deference or refuses to assert its power in some situations.
You're right at the start of building an uncodified constitution on top of the old one just as we did on top of Magna Carta.
It's not entirely new to Trump; every President in my lifetime has pushed on this except maybe Carter. And sometimes they push back (Roe v. Wade was part of this uncodified constitution and probably needed to be a written amendment.)
It could work out but it's important to understand that is what you're doing. And it's not just Unitary Executive theory or presidential immunity; the emergence of the Supreme Court's "shadow docket" is emblematic of the same process.
(Am British expat, btw fwiw)
We had Boris Johnson and Liz Truss; the fact they happened at all demonstrates our system still has vulnerabilities.
I’m no fan of venerating a written constitution either, but I have an anxiety about parliamentary-sovereignty: events like Brexit and BoJo demonstrate a need for something to bind parliament, lest we get a 21st century Oswald Mosley. Right now we have… decorum and monarchy: a firewall made of damp serviettes.
Prior to Brexit, I feel there was some (if vague) kind of accountability from parliament to the EU - which is why/how Brexit revealed the fault-line in the Tories that split the proto-fash types from the amoral-business types to give us Reform/Restore today, and they’re uncomfortably popular. So what I’m saying is that Brexit weakened our accountability mechanisms more than people realize as the national-conversation focused only on economic impact and immigration.
In summary: do not convince yourself that “it cannot happen here”; nor that current safeguards (if any) are sufficient.
Oh I am not convincing myself at all. Like I say, every government ultimately ends up risking it and getting yanked back by the Commons (or the Lords in the case of e.g. 42 day detention of terror suspects, various other bad overconfident schemes like ID cards).
It's a pattern, a cycle. It is always under a sort of measured level of threat. But it is largely stable because at any time the party in power always has an iconoclast (Heseltine, Cook, Davies etc.) who is willing to get in the way, and there is essentially a Commons (or Lords) tradition of making sure that person gets heard.
And because the Prime Minister isn't really that special — just the loudest voice in his party, subject to its back-stabbing — there is a pattern to their failure, too.
(There is a bit of a concern that we're getting too used to them not lasting more than a couple of years, but I do get rather frustrated at the both-sidesing that is including Starmer with a run of utter inadequates from the other side)
Again the point I was making was only that the USA is getting a little further down the line towards wrapping up the clarity of the written Constitution with a layer of yes-but convention... It gets a lot more difficult from here.
There is a narrow window to pull it back, but as much as I think they are for the rule of law, I don't think the Democrats, even with potential rule-of-law-asshole injections like George Conway, are truly prepared to hew that close to the Constitution themselves.
[dead]
Even if Congress may reverse things in the future, there are many opportunistic things happening right now, and it seems like the spacex situation is one of them. The emerging picture feels like a 'flood the zone' strategy (not by coordination, but by practical effect).
While other commenters have pointed out lots of details that point towards the favorable structural environment going forward, another idea that roots my thoughts towards this is that by creating facts on the ground, they are defining the new starting point.
Ultimately, reversing all of the different 'wrongs' or irregularities will be costly to both the opposition's political and attentional capital.
> the current political direction points squarely to Republicans getting absolutely pummelled in the midterms
Even if so, are the Democrats really going to do the house cleaning required to fix this? Their recent history implies that they'll try to pretend things are running normally, until it all explodes in their face (again). Maybe I'm wrong, and they'll actually fight for the country, but... I'm not surprised that companies (and markets) are expecting them to just... not.
Ah to have such hope.
Even if I was this optimistic, the executive with a stuffed supreme court is not going to care what congress thinks.
We'll sooner declare market manipulation a form of speech.
Brother I'm not clear there will be a midterm election and I'm confident most of the electorate on both sides feel very disenfranchised voting and suspect the system is not working on the level.
Going to do my best to respond to this while still following the HN guidelines:
> Please don't use Hacker News for political or ideological battle. It tramples curiosity.
RIMR says:
> nobody will ever challenge this, given the current political direction of the United States
It's obviously hyperbole to say that NOBODY will EVER challenge this, but I'd say it's directionally correct:
1. The Supreme Court is controlled by a conservative, pro-big-business majority that makes it very difficult for any legal attempts to challenge Elon's actions to survive litigation.
2. The United States Senate has a conservative, pro-big-business bias due its over-representation of rural voters and its internal norms (filibuster)
3. The United States House has a conservative, pro-big-business bias due to the gerrymandering efforts of Republican-controlled state legislatures across the country (which the Democrats have tried to counter and failed, see Virginia)
4. The conservative, pro-big-business Supreme Court has ensured that elections in the United States overall have a conservative, pro-big-business bias due to the unfettered spending allowed after Citizens United.
So yes, the winds seems to be against Republicans and Trump in the mid-terms, but the structural biases of the government are still very much pro-big-business, pro-capital, and anti-regulation.
It will take much more than a single mid-term cycle to reverse that trend.
> the current political direction points squarely to Republicans getting absolutely pummelled in the midterms,
I am not this optimistic and I think it is dangerous to be this optimistic. Even putting aside mischief, just as a matter of reality, the gap always closes. A loss, yes. Hopefully. A slap around the face, maybe. Perhaps enough to get a clear win in the House. A pummelling? Nah. I think they might still keep the Senate.
Trump supporters falling out of love for him could well just lead them to focus their attention on down-ticket Republicans who figure out how to make them feel OK about their past choices, and since they need to feel OK, they will come out. Wholesale rejection of the party is unlikely since it is already splitting into two factions.
> 150 of world’s most expensive modern hospitals
https://sfyimby.com/2024/06/construction-underway-for-4-3-bi...
Probably not the most expensive , just first I thought of … so maybe it’s more like 10-15 typical urban hospitals.
is spacex a space company? I thought they were an internet provider that wants to use their strategic advantages to get into AI including AI infra like data centers.
SpaceX since the IPO is an AI company with two side projects: social networks and space.
I say this based on their filing which says that the vast majority of predicted profits will come from their AI company, citing a $36.5T total addressable market.
This is peanuts compared to the 100T TAM in my prospectus around posting insightful comments on HN and Reddit that will help transform humanity.
SpaceX and everything Elon are stock companies - they're Microstrategy but with a veneer of a real business slapped on top.
That’s the real response. It’s all financial engineering
[dead]
SpaceX isn't a space company anymore, it's an AI company. In their IPO filing, of their projected $28T total addressable market, only $370B (~1-2%) of that is from space [1]. The rest is primarily AI, with a sprinkling of Telecom revenue from Starlink.
Given xAI's Grok is way behind ChatGPT & Claude on coding capabilities, whereas Cursor was able to get in spitting distance of them w/ Composer 2.5 by simply running post-training on Kimi K2.5, I'm not sure Elon could dream up a more perfect strategic fit.
Cursor likely has the largest, highest quality dataset of any private firm for training new coding models, which would compete SpaceX's trifecta of becoming a viable competitor in the AI race:
1. Access to compute (they have so much that they're renting capacity to Anthropic & Google)
2. Liquidity for R&D+M&A (largest IPO in history)
3. High quality training data (this Cursor acquisition)
> Isn't it kinda bizarre that Elon is pivoting SpaceX to something else?
In a vacuum, absolutely yes. But in the bizarre context of the AI economics, chaotically scrambling to bring everything you need to compete in-house makes perfect sense.
Arguably, when compared with either OpenAI, Anthropic, or Google, SpaceX/xAI now own the most compute, are the most financially liquid, and (assuming the Cursor acquisition goes through) have the largest corpus of high quality training data.
We may very well be a couple of months away from a Grok release that goes toe-to-toe w/ other American frontier models, IMHO.
So when you look at this as a $60B play to capture an additional 10-20% of an estimated $26T total addressable market, it makes a lot more sense. Now, whether that projected TAM is even remotely close to reality (or even just enough to make Cursor worth $60B) is another question entirely.
[1]: https://www.satellitetoday.com/finance/2026/05/20/spacexs-ip...
* edited to add source for IPO numbers & tweak grammar/formatting
No, it really still makes no sense. Where's the moat around what Cursor provides? If Cursor is really that great, surely something equally great could be developed for a measly $1 billion or so? Is it brand recognition? An established customer base? Surely they don't have $60 billion worth of either.
The key point I think you're missing is Cursor's "moat" isn't around their product or brand, it's around the gigantic corpus of usage data they've almost certainly collected.
It is simply not feasible train an LLM to be as good as these frontier models are without a TON of high-quality examples of what "good" looks like. Every time a Cursor user (who didn't opt out of analytics) does/doesn't hit a "retry" button, or rejects/accepts an LLMs output, it allows Cursor to log record of a specific LLMs output and a binary signal of that output's quality.
Given they've been at this since 2022, and for most of that time sat comfortably at #1 in market share among comparable AI coding tools (only recently getting topped by Claude code), Cursor likely has the largest, highest-quality, SWE-specific dataset in the industry by a sizable margin.
Grok being so late to the party could only train on twitter data in combination with whatever they could source publicly or purchase privately, and likely hasn't had anywhere near the usage they'd need to build up their own competitive dataset from scratch anytime soon.
If you believe (as SpaceX seems to) that the AI's total addressable market is over $26T, and acquiring proprietary, high-quality training data is the difference between capturing ~1-2% of that market and ~10-20%, then $60B starts to look like a bargain.
>they have so much that they're renting capacity to Anthropic & Google
they could not make use of it! That's why they're renting, they don't have access to good distributed training software, so the heterogenous cluster they made was bad for training and they don't have enough demand for inference.
>Arguably, when compared with either OpenAI, Anthropic, or Google, SpaceX/xAI now own the most compute,
lol wut. You think they have more compute than Google?
> are the most financially liquid,
You think they have the most cashflow?
>the largest corpus of high quality training data.
What? You think they have higher quality training data than Anthropic, Google, and OpenAI?
>We may very well be a couple of months away from a Grok release that goes toe-to-toe w/ other American frontier models, IMHO.
Training on what? They rented both of their data centers away!!!
They are building a new one right?
> they could not make use of it!
Grok 4.3 was made on Colossus 1? It's not frontier quality, but it's definitely not nothing. Also, Cursor brings more expertise to help them improve their utilization further.
Additionally, I struggle to see how renting out excess capacity is anything but good. It brings in a ton of cash at a price premium, and ensures they have headroom to gradually phase out rental capacity as their internal demand increases.
> You think they have more compute than Google?
Fair point. No way they have more raw compute, but I do think it'd be fair to say they have more "excess compute capacity" than Google, but even that is pretty speculative.
> You think they have the most cashflow?
Again fair point. I was overestimating how much cash comes from a $2T IPO before actually looking at the numbers. My revised take is that SpaceX/xAI are now in-line with the other labs on cash liquidity, rather than leading (where pre-IPO they were way behind)
> You think they have higher quality training data than Anthropic, Google, and OpenAI?
Yes -- I'm standing by this one. Cursor has a multi-year head start on agentic coding data collection, and a GUI UX that likely provides richer user sentiment & quality signals than something like Claude Code.
Obviously OpenAI & Anthropic have far larger proprietary datasets for chat histories, and Google is an undisputed leader in data hoarding. But when it comes to agentic coding specifically (AI's most compelling use-case, IMO), I think Cursor's data is a HUGE deal. This is backed up by how good Composer 2.5 was given it was essentially Kimi K2.5 + Cursor data.
Additionally, I also suspect its possible to mine Twitter for user engagement & sentiment analysis to create surprisingly useful datasets.
> Training on what? They rented both of their data centers away!!!
I thought they only rented Colossus 1, but you're right they're also renting out a portion of Colossus 2. That said, they still have 2/3 of their Colossus 2 facility available for internal, which is likely enough to build something competitive. They also have 90-day termination agreements in place once they forecast a need for more internal capacity.
All that said, while I appreciate your pushback & I was overly hyperbolic on a few key points in my last comment, I stand by my core theses: The Cursor acquisition (assuming it goes through) combined with the SpaceX IPO puts xAI/Grok right in-line with the other big labs, at least in terms of positioning.
Whether they're able to execute remains to be seen, but I would not at all surprised to see a frontier-quality Grok 5 or Composer 3 release before the end of the year.
Since Cursor is kind of a Select-Your-Model front end for various AI coding systems, how much of it could be to just make sure Grok gets to the top or default choice of that Select-Your-Model list?
Because cursor gets some of the highest quality training data from the world's programmers and responses from the full ecosystem of model vendors and access to active code bases. XAI wants the data.
Highest is extremely subjective in case of cursor. It’s not exactly used by the experienced programmers and caters mainly to neophytes
> caters mainly to neophytes
Perhaps that's where the money and strategy is. (a) stronger need; (b) if you can build systems without real expertise, you don't have to stomach their salaries or politics.
Umm. It’s real development work in real settings with real model output. That is a high quality dataset. The fact that it isn’t good code from elite engineers is confusing what good means in the context of coding agents. First is how to respond to a range of prompts. For that you need diverse real world conversations. Second is the ability to respond with good code. That is about labeling or other data curation after the fact or other training methods. So it’s a downstream consideration
Ah hah, this is it. I was also confused - the tool isn't the thing. It's the behavior analysis capability.
Plus user-funded model distillation lol.
Prediction -- SpaceX rebrands within a decade like Meta or Alphabet. And they turn out to be a AI company not a space company.
How is buying a company sending messages on the Internet for $45b in the owner's interest when Westinghouse, who built nuclear reactors sold for under $10b? The market is irrational.
I think it has something to do with the fact that the first company is better at spreading messages on the Internet than the other at building reactors.
It might not be irrational if there is more revenue in people who are willing to spend money on internet messages than there are on those willing to spend for nuclear reactors.
So far, I see way more companies paying each other in a circular fashion than individuals forking out anything near whatever tam SpaceX is projecting for AI.
when you see the list of major investors in Cursor that will never break even if it stays independent, and compare it to the investors in SpaceX, it all makes sense.
Please don't let Musk know that his money can be used for the real world, mundane stuff.
Because he will buy 150 hospitals and drive them to ruin, private equity style.
We are way better off when he pays abstract amounts of it for abstract stuff to some random nerds and grifters.
His money is not a resource that could be put to any good use. It's a liability for all of us.
Gotta pump that stock price with constant news buzz
Why do you think it's a space company? SpaceX didn't even define ITSELF as a space company in its IPO filing.
it is called SpaceX .
Hardly the fault of anyone is it for not reading a 100+page document meant for investors when it literally in the name .
If they don’t want us to think of them as a Space company they could have taken the xAI name (like how grammarly did with superhuman) or called it Musk Inc or whatever else.
Weak. Would Yamaha list itself as just a musical instrument company?
It’s not real money, there’s no check being written for that amount and then deposited into someone’s bank account.
I mean, the answer is obvious if you do not deliberately try to put a message in the worst light possible:
- "Space company" has a major LLM+datacenter business called X.ai.
- LLM for coding is a big business, as you can see from trillion dollar valuations of Anthropic.
- Cursor is popular and gives you a headstart on the business.
- Instagram was bought for the price of many many hospitals. Uber is more valuable than companies owning the cars. Different business models, entirely different valuation models. Not sure what that comparison entails. You know it. I know it.
Whether it is a good purchase or not, we may not know, but we know your characterization is just outright dismissal without much rationale behind it.
I would agree but in this case they are buying the talent and the data.
Smart move as they will pay in stock, and if stock is overvalued by 2x this means you get 50% discount
xAI is a subsidiary of SpaceX and runs several of the worlds largest compute datacenters.
Giving retail traders a reason to turn their life savings into exit liquidity
yeah but someone -- no one has explained who -- still thinks this is a good idea
Might be a play for talent. Recruiting in this space is hard... and expensive
50 billion worth of talent?
Composer-2.5 is a good model, grok-4 is not.
Its that simple.
XAi is part of SpaceX.
Space exploration requires significant software development.
Take a look at the Apollo 11 movie: There was quite significant computing power for the 1960s putting a person on the moon.
Tesla isn't a car company anymore either.
SpaceX is planning to do data centers in orbit. Engineering issues aside, that might dovetail with Cursor nicely if timings work out.
I cannot believe full grown adult are taking that seriously
As long as the stock goes up, everyone is happy, right?
[dead]
It is crazy that a company like SpaceX was allowed to exist.
The whole big tech industry needs Microsoft/MSN style breakups again.
I bought a small amount (4 shares) of SpaceX stock on IPO day for $160/share. This Cursor purchase does not upset me as a very minor shareholder. Elon Musk seems to have a unique ability to help grow companies that lead to shareholder value. In particular, it seems that xAI overbuilt data centers because their model fell behind. SpaceX could lease data center capacity to Google or other big players (which they're doing), or they can use it internally. Buying Cursor lets them do it internally.
My employer recently switched us all from Cursor to Claude Code. Aside from my personal preference for having a chat window inside VSCode, Claude Code is painfully slow compared to Cursor for my workloads. I think part of this is due to Claude's massive bump in popularity without a similarly rapid build-out of compute. So, the low-hanging fruit for Cursor is to have a massive speed advantage over Claude Code and regain popularity that way. (My current paid AI subscriptions are ChatGPT, Gemini and Cursor. I do not personally pay for Claude.)
And as far as the pivot goes, there seems to be speculation that Elon Musk wishes to roll up all his companies into one big company. So, it doesn't really matter if the AI company lives inside SpaceX or Tesla, since it'll all be one big thing in the future.
Part of Elon Musks strategy seems to sell some kind of hype that does not materialise or at least won't for long (Mars, autonomous Cars) The vast amounts of money collected are then used to develop products that are still a significant progress in its market. Now AI is where all the hype is. It's difficult to sell some hype without AI currently.
> bizarre that Elon is pivoting SpaceX to something else
[dead]
Didn’t you see Moon? They need a Gerdy
SpaceX is just a vibe company. Nothing they are doing makes sense on a valuation basis, which their investors are eventually going to painfully figure out.
I’m worried all this crap will distract and damage the things they are great at (rockets) similar to what happened to Tesla where outside the stock price it’s pretty dismal
Like Tesla the stock is held by many powerful and rich people.
Elon is smart and doesn't buy things alone, he brings in many people so they all have a interest to protect their investment.
SpaceX’s interest is being an Enterprise AI corporation, they identified it as a $24 trillion addressable market, in comparison to their quarter trillion rocket related one
It’s all disclosed in the S-1, you read it right?
In America all you have to do is tell potential investors what you’re doing, its up to the people to use their discretion afterwards
Back in the early days of Heroku (when I worked there), we were all fairly deep into the Ruby community. Ruby has never had a great reputation for performance, but... it seemed like almost a running joke that any time you went down a rabbit-hole trying to understand some weird performance issue you'd eventually discover that @tmm1 had already identified the same issue months earlier, patched it in core, and given an hour long talk about it somewhere. Despite his ability and willingness talk publicly about quite deep technical topics Aman always came across as an incredibly quiet and humble in person. Every Ruby developer has benefited from his attention to finding and fixing performance issues. I'm sure the same can probably said for every GitHub user (where he worked for years).
Congrats to the entire Cursor team! I don't know all of their stories, but I do like to smile and celebrate a little when I see people who are often hidden in the shadows quietly making things x% better for all of millions of us every day for many years getting reward for that effort.
Early Heroku was fabulous. Thank you for your effort!
Yes, it seemed from another word at the time. Now is table stakes. Thanks for showing the way!
Yeah this is really true. It did seem ahead of its time.
Congrats? Usually after acquisition comes layoffs. Especially with Elon at the helm
Especially with Elon? It happened one time (twitter) and if you think twitter was appropriately staffed at 8k employees you need your head read. Cursor operates with a very small team, and practically all have equity, so a $60B exit means congrats are in order.
And Tesla twice https://www.businessinsider.com/tesla-firing-solarcity-emplo..., and https://www.reuters.com/markets/deals/tesla-acquire-parts-in...
I'm not sure what these have to do? Tesla bought an insolvent German auto parts maker and kept all jobs it acquired. This is proof of what?
In staying on track, no I really don't think there will be mass layoffs at Cursor. They have ~300 employees total.
I'm just responding to the assertion is only happened once. No context on why
And the United States Digital Service (USDS) aka the thing they renamed to DOGE.
Every single employee who joined prior to this year will make 8 figures+ off this exit.
Wow. I wonder (genuinely) how such a mammoth payout will affect retention? I guess the option vesting will serve the purpose?
I guess if I found my already-vested shares were worth $4 million, I'd still be motivated to get the remaining $8M by working a couple more years. How surreal that would be, to find yourself extremely "FIRE" (Financially Independent, Retire Early) but also kind of obligated to yourself to keep working hard for a few more years, as walking away from that kind of money seems crazy.
Usually the grants are restructured. Ie instead of what was vested and unvested rank-and-file gets new 4 year grants. And then it works exactly the way you would expect: everyone works for 4 years and then leaves.
>> SpaceX told investors during the IPO process that it sees an addressable market for AI products worth $26 trillion, roughly equivalent to U.S. GDP.
I very much can imagine a future documentary in a few years. With the host asking the audience: "Where were the signs?"
Imagine the East India Trading company advertising the TAM of the new emerging markets in the early 1600's. I bet the numbers compared to the old world would be enormous, and equally unbelievable.
IF, and yeah it's a gigantic really big IF... the future plays out the way people are envisioning it. It's future shock. Astroid mining, gigafabs, star ships launching hourly making space exploration cheap, satellite intelligence, physical AI. The world is going to be a completely different place IF SpaceX and the AI labs are successful. That TAM might be real. It's a literal moonshot, the stuff they're talking about sounds SciFi, but that's why the valuations sound SciFi.
That said, I would not invest anything into SpaceX that I wouldn't be willing to lose, and i personally would not invest until the lockups are free. Moonshots aren't in my risk profile.
This is such terrible history. Please Google things before making analogies.
The East India Companies (Dutch and English) didn't invent or discover the spice trade. They were created to leverage private capital to wrest control of the trade from the Iberian Union, an entity both Company's sovereigns were at war with.
The "TAM of the emerging markets" had already emerged both because of the Silk Route from Pax Mongolica and in the 1500s from the Portuguese Empire's Cartaz trade license monopolies enforced by their naval posts (feitorias) throughout Asia.
The TAM was incredibly apparent, Anglo-Dutch privateering during the war had seized multiple cargo laden ships along with trade route information.
I feel like you are intentially misrepresenting his comment to flex your history knowledge. English privateers captured the Madre de Deus. When they brought it to port, the cargo was worth roughly half the entire national treasury of England at the time. I think that is pretty equivelent to space mining, no?
We've been mining platnium and aluminum for all of humanity, so we know they exist, "The TAM was incredibly apparent". We know there are astroids worth 1 quadrillion dollars. So, if SpaceX has a 1% chance to capture that, thus the valuation.
Good lord.
a) If a single Portuguese treasure ship was laden with cargo of such magnitude, exactly how do you propose this is characterized as a TAM "discovered" by the East India companies?
b) Forget the asteroid, undersea mineral nodes are a total of 250 trillion tons worldwide, with an expected value of 233T USD. This is right here on Earth. Yet no country or company considers these minerals as a TAM, because the P/NAV makes this a net negative. This is the same reason any capex for Texas Shale oil will dry up if WTI falls below 60 for a period of time. No matter how many quintillions worth of oil are left in the shale reserves, if it costs more to extract than it will return, it isn't done. Arbitrary valuations mean nothing if they aren't economically viable.
c) Back to history. "Half the English crown treasury" sounds impressive, but the actual value was an estimated 500k pound sterling, upper bound 200M GBP in today's money. Quite a bit far away from 1 quadrillion. The English Treasury famously passed on a debt of 400k pound sterling to James 1 after the Spanish war, a contributing factor to a crisis between Crown and Parliament relations that eventually helped lead to the English Civil War. Suffice to say, a wartime treasury isn't very large.
Humour me for a bit. Let's assume we need to Hohmann transfer 1 quadrillion worth of minerals (30M metric tons) from a location in the asteroid belt to Earth. The DV necessary is 5000m/s, leading to a total energy calculation of 7.5 x 1e16 J. This is 78M tons of hydrolox fuel. 780k launches assuming 100 tons per launch (Starship). Are you honestly telling me that SpaceX is justified in adding asteroid mining returns to its TAM (1 quadrillion) because you believe it's economically viable to make 780,000 (lower bound) separate launches of fuel payloads from Earth to this asteroid? If ISRU is your claimed solution, then where is the Madre De Deus of this ISRU demonstration?
The most valuable substance in space rn is water. strip hydrogen and oxygen using solar power and place it in orbit. Instead of launching a Mars-bound rocket with 90% of its weight tied up in fuel, you launch it nearly empty, fly to an orbital fuel depot, fill up on asteroid-mined propellant, and go. Big business.
Or maybe you process insitu, nstead of a 30-million-ton rock, you return a compact, high-density capsule of pure platinum. A few tons of pure platinum group metals dropped into a stable Earth trajectory fits comfortably within existing down-mass logistics and could be worth billions.
Thirdly, (prolly the worst of all the options), you place metal slag in orbit and you drop it on peoples heads. Now you've got rods of god that any military will pay billions for.
Edit: Lastly, once you control a vast amount of stratigically important material they would instantly become a sovereign-level geopolitical superpower. The moment a private entity controls a constellation capable of dropping kinetic rods anywhere on Earth within 15 minutes, every terrestrial military becomes fucked overnight. The line between industrial infrastructure and planetary hegemony completely disappears. ya gotta price that in somehow. Then... we really have realised the vision laid out in the book "the soverign individual"
Edit 2: >where is the Madre De Deus of this ISRU demonstration
If everyone could do it, they would have done it. The Madre De Deus was an absolute marvel of engineering at the time. the english knew the portuguese were selling all the stuff from the colonies, but they didn't really know how freakin' profitable it actually was. The Dues was freakin' massive. AKA Starship.
Let’s do a thought experiment. let’s say you find and can mine an asteroid full of rare earth minerals worth a quadrillion dollars, and you then flood the market with your yield. is this asteroid’s yield still worth a quadrillion dollars?
yeah, you hoard it like debiers and use it to build the city of elon, mega factories, or a giant platnium statue of trump to gain politcal favor. You def don't dump it on the market.
But, here's what you'd really do: https://news.ycombinator.com/item?id=48572319
How dare you ~ the OP just now reading your thought experiment.
> We know there are astroids worth 1 quadrillion dollars.
> So, if SpaceX has a 1% chance to capture that, thus the valuation.
Surely 100% chance at capturing 1% of the Veblen Good asteroid is worth incredibly way more than 1% chance at capturing 100% of it.
The TAM that spacex quotes has nothing to do with Space. It is all AI. As in grok. This is a joke, and people that believe it are just providing exit liquidity.
Its not grok at all, their business is renting compute ie. infrastructure. If they have any hope of success its renting out space base compute.
AI hype is full of bad total addressable market hypotheses. There just isn't $26T of money to buy this stuff.
Patrick Mackenzie, at the time at Stripe, about 10 years ago, talked about how Stripe's growth depended on growing the actual internet. Not just growing their market share within it.
Google, Meta, etc, have projects dedicated to getting more people online, because when you have 2bn users, you kinda have to make more users.
If that were to become reality, and I won’t guess at the odds of that, is because it grew the world economic pie. Thinking about it in terms of the size of the pie today is treating it like a zero sum game. It’s a different game.
> There just isn't $26T of money to buy this stuff.
But there is. Like, it is less then GDP of US only
The valuation bubble on AI is threatening the entire global economy at this point.
What do you mean, of course everyone in the first world, including babies and the elderly, is eager to pay Elon some $20k/year.
The GDP of the entire world is $126T. Each and every year. And, thanks to USD inflation, that figure is increasing rapidly.
Based on the disclosed financials by SpaceX, they have made $48B in revenue since 2023 with a cumulative loss of -$41.3B.
And yet, based on stock price and market cap, they are worth about as much as Microsoft. Microsoft generated $80B of revenue and $31B in profits...per quarter last year. SpaceX will never, ever, generate $124B+ in annual profits.
What all the SpaceX naysayers don’t understand is that they effectively have a monopoly on selling drugs to aliens. Think about it, man.
Until the alians realize this and throw the drugs in the harbor, leading the a US vs alians war to open them up for free markets access. Not sure this can actually realize their ambition.
What about the compute deals? $3.1B a month for half their compute? Elon built $6B of rev in 122 days. What can he do now that he's got a $85B war chest of cash?
Buying NVidia gear early and renting it out isn't exactly business model. Like what is their moat? While literally everybody is bidding up hardware?
The only way you can see this continue as a model if space based data-centers actually can beat everything land based. And that is very questionable.
> What can he do now that he's got a $85B war chest of cash?
Increase the profit of Micron and NVidia?
>Buying NVidia gear early
xAI was retardedly late to the party. When Elon Musk founded xAI in July 2023, competitors had already spent years buying up massive silicon footprints and locking down cloud partnerships.
OpenAI had a 3 year lead, anthropic 2.5 years, and google was building TPU's since 2015... what are you talking about?
First, many more people then just them buy these things and have ramped up their spending over the last 2 years.
And others if they were earlier, the explosion and demand of investment increased exponentially over the last 2 years, as we can see in the profit margins of the memory companies for example.
My claim was not that they were the first ever, but that Musk mobilized a huge amount of resources fairly quickly in a time where prices were exploding and before many others.
Early enough that compared to now he got a ok price and he can rent it more expensive then he bought it.
He can do things and a scale and timeline that is impossible for other to achieve and why he can make so much money? I agree.
I'm not saying its not a reasonable business, but we are not asking 'can it be a business', we are asking can it be one of the most insane companies in world history. The suppliers have realized whats going on and they are racking in insane profit.
That's why all the Musk fans have realized that they need to scream 'Terafab'. But that is a fanciful concept and unprecedented in history of semiconductors.
>we are asking can it be one of the most insane companies in world history.
"So we went from, he just got lucky cause he got in early", to, "okay, he wasn't early he was just really fast", to, "the suppliers now see he's making a ton of money so they are going to raise prices" to, "terafab is impossible", QED?
You seem to misunderstand how tier 1 capital allocation actually works. You don't get fucked by your suppliers when you are adding $B's to their bottom line. XaI never "Stopped" purchasing GPU's in 2023. They signed multi-year long deals for as much hardware as nvidia, supermicro, dell and micron/skhynix can provide. They are the last to get squeezed. They never stopped adding on to Collosus 2, and they are actively building collosus 3.
Lastly, nothing is "Fanciful" about terafab. ASML makes the machines, Intel provides the cleanroom tech, and 14a pipeline... these are known variables that need to be tackled. Tesla and SpaceX are world renowned, high throughput manufacturing facilities with the ability to make lines more profitable than their competitors. .. no one is claiming he's building a TSMC-style silicon foundry. They are building a massive compute fabrication/clustering footprint. It is a formalized $55B industrial plan filed in Texas, backed by a hard partnerships.
They have quite short cancellation notices and there is a question why do you need to rent it if you are an ai company? How is it grok not running the world using that compute.. And possible cancellation will erase a large income stream, think when gpus get old or sooner when Google does not need it anymore.
>there is a question why do you need to rent it if you are an ai company
Uhhh... $3.1B a month for $20b in capex sound like the best investment ever?
The compute deals are going to go the same way as all those data centers that openai said they would need to build.
The compute deal is for their depreciating hardware.
Matter of time before its no longer considered good enough by the main players.
Anthropic are building their own Datacenters so it's a stop gap
an RTX 3090 or older Quadro/Tesla enterprise cards are still selling/ renting... 5 year depreciation is only for tax purposes.
Anthropic is not building their own datacenters... They are mainly giving equity for colocation with AWS. I cant find one example of them ground breaking on anything... Can you give me a source?
They are building one in uk
https://www.themorningintelligence.uk/anthropic-lines-up-tee...
yeah, and power is not availabe at capacity until late 2029? They haven't even signed anything, the 'deal' is in principle and BP has been actively trying to challenge it because they want the same power for a hydrogen facility.
So, to circle back to my original point, Anthropic has not broken ground on actively building any datacenters.
Maybe they can monetize and trade AI tokens like Enron was going to trade bandwidth?
Enron was never audacious enough betting every US man, woman, and child will spend $28k/year on their generally nonprofitable business with one exception -- Starlink.
Patrick Boyle covered the SPCX trajectory fairly well... =3
https://www.youtube.com/watch?v=wKXgeNwNRJ4
This feels very 1999.
Market maker for AI tokens. Now that’s an original idea!
Well the TAM can match GDP if GDP evaporates!
That's the number AI boosters are spreading too "The TAM for AI is all of humanity - this includes every person and every company. So, imagine this huge pile of future revenue." I agree that TAM is likely huge but is SpaceX most suitable to capture that TAM? Unlikely. But for now everyone wants in on the AI hype train and FOMO of losing out one any company in the AI space.
In the long term most markets are duopoly with small competitors. And personally I see OpenAI and Anthropic duking it out rather than SpaceX.
I know this is just nonsense wish thinking, but apparently the investors disagree and I have zero clue when they will also stop giving Musk their money.
I think the bump since IPO can be explained at least partially by low float not meeting demand. I've seen a lot of accounts from retail investors who entered the lotteries saying they only got a small fraction of what they wanted, hence demand is kept artificially high. Probably intentionally, since it essentially allows the optimists to dictate the price.
(edit: This is not at all unique to spacex, of course, but given the nature of Musk's companies and their "fans" it's logical that they would employ this strategy. They are also doing a staggered unlock to avoid upsetting the market when insiders start dumping their shares.)
"investors" really over-sells who these people are. They're people with gambling apps on their phones who think SpaceX is great because of Musk's politics.
Signal and noise. Lots of noise. SO much noise.
To be clear, I don't know which part is signal and which part is noise any better than anyone else.
This is a stupid comparison, but Mojang/Minecraft was acquired for 2.5 billion in 2014.
Arguably the most popular video game of all time, which has brought joy to hundreds of millions of people for years and years, was valued at 1/20th of an AI startup that will soon disappear into irrelevance.
While Minecraft is just a game, I'd argue it has more societal value than Cursor. The way things are valued is nonsensical to me.
Will Jack Black star in the Cursor movie? Only time will tell.
> While Minecraft is just a game, I'd argue it has more societal value than Cursor. The way things are valued is nonsensical to me.
Well it's because societal value is not profitability. Only question that matters is if Cursor can wind up worth more than 60b. Not even in raw revenue so much as ability to keep shilling the same story.
For comparison couple of months ago I acquired zed, pi, opencode etc. for $0.
I think they eye the models more than the agent or IDE
The models are the least valuable part of all of this.
I'm not sure whether Elon's interests are in the model. Maybe the existing distribution channel which is cursor's users? The whole ai business models seems tough to comprehend. Huge amounts being passed from one player to the other.
If Elon is smart about it, he will use the still relevant enterprise-friendly brand "Cursor" to get a foot in the door.
If he is dumb about it, like he was with Twitter, he will just rename it "X" and delete any value the acquisition had.
Minecraft erases productivity. AI coding vastly increases it.
The economic value of a tool is in the economic value it can produce. The economic value that Minecraft can produce is limited to how much money 8 year olds can wheedle out of mom and dad. The economic value that Cursor can produce is probably in the trillions, if they play their cards right.
You're also forgetting that money is worth a lot less today than it was in 2014. $60B today is $42B in 2014 money. Still 17x that of Mojang, but not the 24x that it appears at first glance.
AI coding vastly increases the amount of code produced. It does almost nothing for productivity.
> While Minecraft is just a game
It's more than just a game. It's a bespoke social network. It's a merchandise generator. It produced an entire hollywood movie. It's become a cultural reference.
Do kids grow out of it by a certain age though?
It's been around for about 17 years, and if anything is more popular now than ever. So even if lots of kids grow out of it, there are new ones being born every minute.
I imagine at least part of the reason this doesn't make sense to you is because you're not a decision maker working at that scale. It made sense to someone at a high level. The calculations that person is doing is likely different than yours.
This is also why it's difficult to make money selling to consumers. They run different calculations than enterprise buyers.
There’s quite a few historical cases of mergers acquisitions being poorly valued aol time warner for example. So maybe v the calculus is different from the execs perspective but it’s not necessarily any better.
things which help other people make money are overvalued relative tothings which are valued intrisically
Elon has all his datacenters idling, and his xai failed. What else would you buy to still stay in the race?
Idling? Where do you get your news? He's leased a ton of capacity to everybody else. Makes little difference to Musk whether Grok is using them or someone else is paying him to use them. We all know from AWS's profitability that renting servers to other people is great business. You get a check every month (often from other people's VC money) as opposed to having to figure out how to monetize.
Reselling raw GPU power means your competitors are able to extract more value out of it than you can.
If that is not a failure for an AI lab, then I don't know what is.
I see it more like McDonald's can't produce burgers fast enough because of a meat shortage, but Burger King has their own supply chain that allows them to supply extra meat. McDonalds now needs to buy patties, possibly at a higher price than they usually pay, from BK just to be able to meet their burger demand.
Is BK in a bad position? They get to run their beef plant at 100% utilization regardless of how good they're doing at marketing hamburgers, and they can use that money to keep improving their hamburger quality or marketing.
That's a weird analogy and also not fitting at all.
Let's stay in the similar space and talk what it's really about:
Imagine google builds 2 data centers and sells the raw compute(not services!) to AWS so that AWS can sell their service on top.
What an embarrassement that would be.
Right, value != price. Now, is this a bug or a feature? Discuss (serious question)
My wonder is if anybody would buy anything if value == price.
The value of water to a person dying from dehydration is infinite compared to someone who's adequately hydrated. By that logic, this bottle of water is worth every possession, good thing, in the present and in perpetuity because that is the opportunity cost. Nobody would choose $50 million in cash to reject the water because the money only has value if you take the water.
But hypothetically, if the value and the price could be finitely defined, a "fair trade": let's say 10 apples and 1 watermelon are each worth 10 utility units. Price still can't equal the value. We don't eat utility units. The watermelon inherently provides a different value than the apples. An apple isn't a substitute for a watermelon.
I think this is my long winded way to conclude that trying to compare everything is apples and oranges, but somehow we still try to give it a dollar amount.
> The value of water to a person dying from dehydration is infinite compared to someone who's adequately hydrated.
But most people are adequately hydrated in steady state, and to them the value of water is not that high. Since today it is necessary to sell an item for the same price to everyone, selling water for infinite money would lose nearly all customers.
(Not to mention that it’s possible to acquire water from natural sources, for free)
minecraft is the actual metaverse.
how much did facebook spend trying to recreate it?
Somewhere between ~15-75+ billion USD. Higher end if you want to include the hardware R&D/etc, which is arguably core to their metaverse concept. I love the idea that they don't even have a product _called_ metaverse. Most of the VR content developer acquisitions costs by Occulus Labs are undisclosed.
https://www.fool.com/investing/2024/04/01/meta-platforms-has... https://en.wikipedia.org/wiki/Reality_Labs#Acquisition_by_Fa... ~1.6 billion
World of Warcraft is the closest thing to the metaverse. But I'm biased since it's my favorite game (TBC, WoTLK, etc)
u cant compare productivity to games on the metric of bring joy - ofc productivity tool isn't bring joy like a game does.
Happy and fulfilled people are generally more productive, indirectly.
Minecraft is just limited for additional value creation, i.e. it's not a tool for creating more value
> "In its IPO filing, the company had said Cursor's access to developers' data, including coding requests and design decisions, could help improve its AI models such as Grok."
They're all stealing your IP and selling it back to your competitors in the form of tokens.
Yes, that is in fact how models get better at coding.
Such a ridiculous stance: "I want LLMs to code for me, but I want them to be trained on other people's code, not mine, duh".
> "I want LLMs to code for me, but I want them to be trained on other people's code, not mine, duh".
Who ever said that? Have you actually heard that from your fellow programmers in real life?
If the code I wrote actually made even the slightest discernible difference in LLMs I'd be so honored. But it won't happen, as it's just 0.00001% of all the training data.
Real life? Most not. Hacker News? Absolutely. Literally the comment I am replying to.
> But it won't happen, as it's just 0.00001% of all the training data.
Are you familiar with Tragedy of the Commons?
Tragedy of the Commons is just an analogy - so not the fact.
Tragedy of the Commons is not an analogy, it is an inevitable result of a large fraction of participants in a coordination game defecting due to perceived individual advantage.
Sounds good? They can pay for code they want to train on. There are plenty of companies sending me offers to code training materials for them for $50-100/hr. Don’t expect to charge me an arm and a leg for inference and then also train on my code.
There are already opt out buttons for training in Cursor and Claude Code… if you don’t want it then turn it off. If it was worth enough money to them they would offer a monetary incentive like discounts but none of them have yet
They are talking about the millions of lines of code they stole to make the product in the first place and I'm sure you know that.
This just makes the inference more expensive for you?
How about: "I do not want it to code for me or anyone if it steals from someone."
Interesting how our generation which grew up using Napster now has so many intellectual property extremists. By this logic, even humming a tune you heard on the radio is theft.
Ok, now that you mentioned it, I actually want that.
Who are you quoting?
It's a common sentiment. An example from few hours ago: https://news.ycombinator.com/item?id=48558954
> I have absolutely zero interest in free. I honestly don't think I'm even remotely in the same demographic as people using free tiers / models. I want to pay. I don't want my data used for training...
They want to use LLMs trained on others code but don't want to contribute with their own.
Not casting judgement, just pointing out.
It makes sense from a business perspective-SaaS firms value the ability of coding agents to accelerate development, but also worry the models will learn the secret sauce of their business and destroy its moat. So their desire to contractually exclude training on their data has some logic to it.
(Disclaimer: Not speaking for or about my current employer, just a general industry observation.)
I don't really use LLMs myself, but if someone wants to have any kind of software business then having the models trained on their products isn't ideal.
I mean, AlphaZero et Al start from zero. I learned writing my own code except for documentation and some textbooks.
Fair point, an AlphaZero of code would be very interesting indeed.
This is the correct take
No one is stealing anything here.
Cursor users are willfully providing it by using their product. Not unlike uploading a personal photo or video to social media -- that's not yours anymore. You gave all rights away when you put it on their servers.
> Not unlike uploading a personal photo or video to social media -- that's not yours anymore. You gave all rights away when you put it on their servers.
99% of users are unaware of that, so it’s false to say they "are willfully providing it".
Do you think that everyone using it and their employers are aware that they are giving their competition the ability to copy straight from their codebase when they ask it to replicate their product?
most serious employers have data retention agreements.
You trust those?
presumably they are subject to audits. i work at a tech company that has to deal with data compliance and we take it incredibly seriously and have automated systems tracking, tagging, and deleting data everywhere.
I have worked at several tech companies that deal with data compliance. Most audits are jokes and most of your data is not secure.
> Not unlike uploading a personal photo or video to social media
I've granted them a limited license to use it.
> that's not yours anymore.
Not by any definition in the contract or in law is this true.
> You gave all rights away when you put it on their servers.
I gave away some rights. I also got something in return. Attention. And at the end of the day I'm completely entitled to turn around and sell copies of this work for profit. The only thing I can't do is sell an /exclusive/ license because that is no longer available.
None of this provides any implication for people who upload code to their own websites. Which these rapacious LLMs bots happily index, sometimes to the extent they actually crush the site, or create unusual costs for the owner.
Finally none of these LLM companies tell you where the source came from. Whether it is copyrighted, whether ownership rights are retained, or whether the code can be used publicly or not, and if so, which license it's covered by.
You're using the lens of social media contracts to understand something far larger and more important. It's lead you to some bizarre conclusions and huge oversites.
What if someone steals my work and then uploads it to facebook and claims it as their own? Do the rights no longer exist because it got uploaded to Meta?
I work at a 20k-employee company with footprint in tech. A couple of weeks ago a decision was made to not renew Cursor license. This week it was made public and has been greeted with groans. Some employees have been publicly saying how ineffective and unproductive they are going to become due to this decision- which actually strikes me as really naive.
Personally I like to use a stable IDE. I used Cursor for a couple of days and then went back to VS Code, largely due to Cursor pushing agentic first approach with V3 update.
Most people are not good at their jobs and IDEs like Cursor let bad SWEs feel more productive because they are committing more code.
Indeed, according to Cursor leaderboard there were folks committing 200k-300k lines of code every month. And here I am intentionally slowing myself down manually approving every AI generated code change and terminal command.
> They're all stealing your IP and selling it back to your competitors in the form of tokens.
The users of those tools are stealing too. The model is trained on free software licensed under specific terms and the output of a prompt will strip those licenses and their terms.
Quit posting trashing. This isn't theft.
It's not theft in the same way a con artist convinces you to give him all your money. The people using it or their employers just don't realize any competitor will be able to ask the LLM to replicate their product and it will copy the codebase they uploaded to them.
This isn’t how training works.
I think he’s saying let’s say I’m your competitor and I need your code and you happen to use the same llm they will provide me a copy of your code.
Yeah, which hasn't been demonstrated in practice at all.
It has. LLMs being able to one-shot any higher-order complexity is entirely dependent on it already being in the training data.
If it has been demonstrated, you're welcome to cite a paper or even study of any kind. Right now, you're just guessing. (I head an AI lab, FYI)
You could just test it for yourself if you're actually interested in finding out (assuming you are a decent programmer, but your bio suggests you have no technical expertise in the area).
I was working in machine learning before you were old enough to take programming classes.
https://github.com/jleclanche
As I said, I founded and run a AI lab. Now get over yourself.
First I never made any claims about myself or appealed to my authority, that was you. I still don't see any evidence of any technical expertise in ML or in programming anything more complicated than python scripts. Maybe you weren't aware but often on the internet people make false claims about their expertise on something. There's thousands of "Visionary leader driving ethical AI innovation and strategic growth" CEOs that know next to nothing about the actual technology.
Anyway like I said you can just test it out yourself and find out that I'm correct. Every skilled programmer already knows this and can predict what kind of complexity an LLM won't be able to handle. And anyone working on LLMs should know that they are completely dependent on their training data. The entire scaling hypothesis was based on this.
There's open source ML&DL projects dating back 12 years on my github, I don't know what you're on about.
Regardless, your claim is "LLMs being able to one-shot any higher-order complexity is entirely dependent on it already being in the training data." which is currently unfalsifiable, and known by every AI researcher to be most likely wrong. You've also said it's "been demonstrated". So stop wasting people's time and link to the demonstration instead of hand-waving a "do it yourself you're smart".
Good lord, it's like talking to a bloody climate change denier, I swear.
I have neural net applications I programmed entirely myself from 16 years ago and all I see from you are some python scripts so I guess I win this stupid game I wasn't even trying to play.
You haven't made a single technical point outside your bloviated claims to authority. I can safely assume you're a fraud because this is exactly how frauds speak. Actual scientists and engineers don't argue from authority they go and test the hypothesis for themselves, the fact that you balk at my suggestion to do this is amusing.
I said you can do a basic test because this is the best way to see it directly for yourself. It's very easy to do especially for an eminent machine learning visionary leader as yourself. You ask the LLM to produce two apps of similar complexity and technical challenge from a software perspective, one that is already in its training data and one that isn't and see which its more successful at. This isn't some controversial take, nor is it "unfalsifiable".
There's also hundreds of benchmarks demonstrating where the limitations are for LLMs, or you can study the progression of LLMs in mathematics and where the gains have been made and see that this also agrees with me. You can watch Chris Hay's videos demonstrating exactly how LLMs perform math layer by layer. Why is everyone using LLMs for search? Because it's an extremely efficient compression of all its training data. Did they figure out the Studio Ghibli art-style all on their own spontaneausly? No, they were trained on Studio Ghibli content. There's so many ways to come to this conclusion. But you seem to be too busy sniffing your own farts to be interested in learning anything about the field though.
This pointless test you're suggesting is a few countries removed from your initial suggestion, which was as a reminder: "The people using it or their employers just don't realize any competitor will be able to ask the LLM to replicate their product and it will copy the codebase they uploaded to them" (which, yes, is unfalsifiable given the search space)
I mean, are you seriously trying to back off that original ridiculous claim into a "code in the training data is more likely to appear in the output than code that isn't"? And I'm the fraud? As I said before, get over yourself.
Yes a test showing an LLM can reproduce an app in its training data and not an equivalent complexity app that is not in its training data is equivalent to proving the statement I made that having your codebase trained on will allow a competitor to copy your product.
And yes, you are obviously a fraud.
My whole team was on cursor for a few months. I enjoyed using it and thought it was the most complete of the agentic coding tools I tried. The thing that got me was the cost. I was switching between Opus and GPT 5.x and was spending anywhere between $500-1000/month. I was using a relatively normal workflow, paste in ticket, plan, execute with dumb sub agents, have the ai test and competing model to validate. The business got uncomfortable with the cost when everyone started doing the same so they switched us to Claude code since it has better cost controls. So far it looks like we won't even touch the $100/month plan and some people would be ok on the $20 plan. Anthropics usage limits is a consistent source of complaint on here but I've found them to be moderately generous in comparison to cursor. Cursor also charges a $.25Mtok premium for 'routing' no matter what model you choose. 5% increase for frontier models but when you're using haiku on sub agents that's a 50% cost increase. Composer is solid but if you don't have deep pockets it's the only feasible model on their platform because of how they bill it. Being an all in one editor/agent is nice but if you're in a language like c# or Java you're already swapping back and forth with a real IDE anyway.
Our experience has been the direct opposite. The way to get the most out of cursor was to leave it on auto - we saw it average around 3.9 cents per request under the old contract (per-seat pricing) and more like 39 cents per request under the new (single pooled cost for everyone). Composer came in a little higher, more like 50c/r, while the claude models were up into the dollars. Meanwhile, if you use Claude-the-app, there is _no_ cheap model and the default switched to Opus in April, resulting in increased costs across the board.
We have both Claude and Cursor here, as well as agents running GPT, things in AWS Bedrock, etc and its my team handle the bills...when people use Cursor on auto, costs are under control, but there's always a dozen or so whale users who'll switch models manually and blow through the budget like it's not there.
Another thing: "better cost controls". There was for example no way for us to disable Fable in Claude, but we could in Cursor. Again, the opposite experience.
For the kind of ML work I do Cursor's "Auto" was too unreliable. It would frequently try to solve relatively difficult tasks using Composer, when it should have been routing to a true frontier model. Annoyingly, doesn't tell you what model it's decided your task deserves, so I often wasted time working through a problem with it, only to realise I was on a dumb model. Then I revert all the work, switch to an expensive frontier model and pay expensive API prices to actually solve the task.
All that is to say, model selection is the main control we have over quality. Giving it up in the name of cost saving will bite you in the long run, especially when Claude Code still has such good plan pricing.
I also had a problem with Cursor auto switching on fast mode which I would never use otherwise.
Cursor really needs a stronger model than Composer 2 which is the root of the problem. Hopefully they will put out something better now that they have better datacenter access with SpaceX.
Yes that was very annoying considering the 5x cost increase for fast. I can wait it's fast enough.
Well, auto can produce garbage code. I'd rather be able to select the model manually myself and so far it's much more economical to give people Claude Code and Codex subscriptions as those are subsidized per seat plans.
The price you're objecting to is the heavily subsidized cost, it'll be 4x that in 12 months
> I was switching between Opus and GPT 5.x and was spending anywhere between $500-1000/month.
I switched to Codex because Cursor was costing me $10k / month. If the price goes up by 4x I'm not sure the value will be there for a lot of people.
More precisely -> the competitors heavily subsidize which causes cursor to feel expensive.
> Recently, we purchased one of each Anthropic/OpenAI subscription plan and randomly ran long horizon coding tasks until we exhausted the weekly limit. It's widely believed that a $200/month plan maxes out at ~$2000/month worth of tokens (assuming API pricing). However, we found that the subscriptions are actually far more generous. (2/4)
https://x.com/SemiAnalysis_/status/2064815044085318040/
Same I’d pay for Cursor again if they get more price competitive with Claude max
Is anyone on HN still actually using Cursor in 2026?
Everyone I've spoken with is now using either Claude Code or Codex (or Copilot because their companies force them to).
Just setup my team with Cursor. They actually got back to me on an enterprise plan (Claude keeps ignoring me). Cloud Agents have been great for keeping multiple streams going at the same time. Adding in computer use has been great for actually testing out features and showing they work for PRs. Bugbot so far has been the best AI reviewer I've tested. Composer 2.5 is great, though still using Opus for planning.
I can do most of this with Claude Code, but there's definitely a cost in maintaining it for the whole team.
Yeah, I've tried many scanners and Cursor Bugbot is easily the best
It's my primary. Claude Code for personal stuff on the weekends. I really just prefer the GUI of having the changes easily highlighted. If I can get something to apply that with Claude Code or Codex or OpenCode or whatever, I'd swap over without thinking.
Same here - their UI/UX seems to serve my workflows/habits the best. And it's strange that no one else seems to be delivering a compatible experience for me. I'd prefer to move away from Cursor after this acquisition.
The Claude Code extension in VSCode is similar enough. I used to use Cursor, but now I use VSCode+CC to take advantage of the better pricing CC offers for always-on frontier. I found Cursor's affordable "Auto" option unreliable, often wasting my time with dumb models.
As in the changes its made? You can do that on Claude Code but by having it do it on a worktree and just checking out that worktree. Then you can see the change file(s) in your preferred IDE
Wouldn’t Kiro fit the bill?
with cursor it still feels like coding, all the others are diff checks it seems like.
I stopped using Cursor because of how terribly optimised it is (worse than VSCode despite being a fork). It would routinely take up 50% of the CPU resources on my MacBook M4 and gigabytes of RAM for absolutely no reason.
I switched to Zed, and I'm never going back to Electron/non-native IDEs.
Yes, I use it as my daily driver at Discourse, the Cursor Tab autocompletion is still the best AI-based editor autocomplete I have used. I switched to nvim earlier in the year because there was no real great AI integrations that didn't feel cumbersome, Cursor in vim mode is a much nicer experience, I like adding lines of code to agent context and so on. All of the extensions from VSCode, LSP integration, and so on are really nice.
That being said I do a lot of work in Codex or Claude now (they all feel pretty much the same to me), and use Claude for manual code writing and tweaks that I feel would be unnecessary for agents to do, or just when I am writing code that I enjoy writing, typically when exploring a new problem I'm interested in, not everything needs to be done at 1000% speed by a robot.
Though I will likely reconsider this now that Elon will own it, either moving to VSCode or to Zed long-term.
Still do. Composer 2.5 is a beast. But even with Opus (and Fable for a few days) their harness is many times faster. The main reason for me to use CC is the $200 subsidized pro max plan.
Also their computer use in the cloud agents (when it works) is a game changer. No need to keep your laptop open / get a Mac mini if it runs in the cloud.
I was on Claude Code the past year, now I use chinese models, but I've used Cursor and they have an ok pricing offering today because of their mix of sota models with usage based pricing along with their Kimi based Composer model with generous limits. I think it makes a lot of sense for the enterprise market, which is the real moat, and not the capabilities/features of the forked ide or app/tui/github bot anyone can come up with today.
I much prefer Cursor to Claude Code. Claude Code hides too much of what it's doing from me (can't monitor the thinking output, can't see output from in-progress commands being run.
I don't even use the IDE -- just the Agent Window interface. I also really like Composer 2.5.
to be fair "nobody" is using grok either
I use grok every day as my go to search engine replacement.
That's good for you but the demand is so low SpaceX is leasing a considerable amount of its compute to Google and Anthropic to compensate. Regardless of whether it is useful, it is clearly not popular
Same Grok is an amazing search engine replacement
I use Cursor for work, but claude code for personal development. I think Cursor is still useful but the most value really is access to latest models
> Is anyone on HN still actually using Cursor in 2026?
Which makes this a godsend for Cursor.
Even if SpaceX stock plummets 90% before the lockup ends, I assume everyone will make out better than any other exit that could have occurred.
Who would buy them that would seem rational?
I switched back to VSCode with Codex and Claude extensions. Just more stable.
I use Cursor for coding. I like to review the changes via the UI. Plan mode is also really strong in Cursor. It bugs me less about needing to search through files and basic coding tasks. I find it also saves the company a ton of money compared to Claude, Claude burns through tokens with no regard.
I typically use Claude for interacting with MCPs and skills to operate on live systems.
My company gave me cursor license after I had already been using Codex CLI for months and VS Code for a decade.
I had absolutely no interest in their VS Code fork. The Agent Window was okay but buggy (eg wouldn’t load branches on Ubuntu via WSL2).
Overall used it a couple of times but still use Codex CLI as my main driver. Might try CC in the future esp. if they unban Fable.
Plenty of enterprises are still using Cursor, though they are facing plenty of pressure because Anthropic and OpenAI bundle Claude Code and Codex which can make it hard to justify an additional license for a third-party harness (why spend that money there when you can buy the underlying tokens instead).
It's still my primary at work (data engineering/platform engineering), running a mix of GPT-5.5 and Composer 2.5. We also have Claude Code subscriptions. I find myself preferring Cursor for most tasks.
70% of the time I use AI agents on a pretty tight leash. I often reject edits and ask it to change things. The IDE integration is really efficient for this workflow compared to Claude (yes, I've tried the Claude extensions).
Autocomplete is still the best available (I've tried both Copilot and Zed); though admittedly it's not as important as it was circa last year.
For the 30% simpler or very well-specced tasks their cloud agents are last I compared way better than Claude Code's/Codex's version of the concept. @cursor for quick fixes in Slack works quite well. Don't get me wrong, it's still quite under-documented, but the others are worse. The integrations with linear, Slack and github are well executed
Composer 2.5 is really fast in their harness at code search/explanation/Q&A tasks (much faster than Sonnet/Opus). It's also really good at debugging, very proactive compared to other models in the same size/prices class IMO. Just due to the speed I actually prefer it to almost any other model for these tasks. I suspect at least some of this may be due to the harness and good codebase indexing.
I don't know why people are down on the Cursor harness. It's good. The main advantage of Claude Code/Codex are the token subsidies; but according to their dashboard I am costing my employer between $100-200/month on Cursor, so the overall price is comparable and only narrowing now that Anthropic is switching many enterprises to API usage.
I also don't understand the people complaining about VSCode bloat and in the same sentence praising Claude Code. Claude Code often uses MORE RAM than Cursor, has a super unstable UI (on my home machine there is input lag when typing ANYTHING in Claude Code) and the desktop app version of CC barely works. The Codex TUI is genuinely nice and snappy, on the other hand.
I'm still on their old 500 requests/month plan and the value is simply unbeatable for $20. I've been able to use agents without worrying about usage for my job and personal projects paired with the $20 codex sub and I dread the day when they finally get rid of the requests plan.
Both teams I run use Cursor as their daily drivers. It's fantastic. I am annoyed with this acquisition.
Yes, let the enshittification begin.
I use Cursor with the Claude code plugin because Cursors autocomplete is really good and I like the way Cursor is set up. But it’s definitely a UX downgrade using a plugin instead of the builtin Cursor agent stuff which I do miss. Claude with Max are just a safer bet economically these days than a cursor plan.
I except that model and pricing gap to narrow over time though
I've recently switched off those to Cursor. More flexible tool, better interface.
I can only stand to use the cursor CLI at this point, but it's usable I like to be able to switch models when one gets stupid or guard-railed, on the pro plan you get so much usage in the composer 2.5 pool you could just use that your whole sub if you wanted.
Zed with Claude code is the best of both worlds
My biggest worry is that Zed gets acquired.
Development is moving away from the IDE to agentic long running workers. I've been using their SDK in this mode - which then forces you to use cursor as model provider. I use a mix of harnesses for different types of agentic tasks and Cursor gets the best results.
I don’t see how Cursor differs in that regard. They have an agent management system since Cursor 3 and some feature that offloads agents to the background or cloud which I haven’t used yet. I always just keep chat windows open if it’s running because I’m always watching that stuff closely.
Seems more like an advanced/niche feature for people who go really hard into LLMs IMO
Earlier this year I had used it because I would rather have a IDE-like exp and be able to actually look at the code. However, recently switched to using claude code VS code extension and it's basically the same thing (plus at Amazon we can only use Claude Code)
Fully on Cursor at work and I love it over CC, OpenCode and Pi that I use for personal work.
I use Cursor, but funny enough it's 98% just using the codex plugin - I kept cursor around on the grand fathered $20 / 500 requests plan, if they un-grandfather me or things change too much I'll zip over to vscode.
I have seen a few codebases lately with AI-bullish teams. Code produced by Cursor reeks of low quality. I’ve tried it but never got hooked.
AFAIK their market is pseudo-technical people who haven’t found the terminal yet.
I use Cursor and it’s been fine. I write a lot of code manually too, so I liked the tight integration with VSCode, my daily driver for about a decade. I used to use Vim, so I’ve “discovered the terminal” a long time ago.
The people steering the agents are the ones producing low quality code. I see little correlation outside of that.
I don’t understand this take. In my company a lot of the Claude Coders seem to be very uninterested or unaware of the code they are producing, while I in Cursor usually click ”Keep”/”Undo” on specific code blocks (with little edits) or sometimes the whole file at once if it’s a low risk part of the codebase. I fail to see how this workflow produces inferior code vs shooting in the blind and maybe skimming a huge diff in one go.
I’m running Opus through Cursor like I would in Claude code, so make sure you’re not comparing different models.
But the core difference for me is I can easily see the code changing in cursor, but not in Claude code. Bigger tasks, I’ll have to drop into other tools to see what Claude is doing along the way. I don’t like that. I like everything being in the same spot
Cursor became obsolete pretty quickly as Claude has improved. Good on them to find an out before they collapsed. $60 billion is a huge overpay.
Yes. Do the heavy lifting in Claude code , Codex.
Basic tasks in cursor. It's decent and damn fast.
All my team members also use it.
i use literally all of them, they're all pretty good. codex is the lightest on my machine which is nice but less featureful - the others are all pretty at parity (minus a few CC-only bells & whistles)
I used their agent view yesterday and the file tree does not update when you add new files.
We use the bugbot. Best code review agent we've seen.
I use it because their pro plan is free for students
> Is anyone on HN still actually using Cursor in 2026?
Most people on HN wouldn't be satisfied using anything. Therefore, probably no
if you use hn you probably don't speak to actual people tho
It's my daily
I use Cursor every day.
Yes, unlike Claude it has excellent response rate and i can leverage their speedy models
$60B is a huge price, but buying Cursor gives Musk something xAI has struggled to build: a popular coding product with real developer and enterprise adoption. It may be the fastest way to catch up in AI coding. The real question is whether SpaceX ownership improves Cursor or drives its users away.
> The real question is whether SpaceX ownership improves Cursor or drives its users away.
I know I'll sound hyperbolic but I'm deeply skeptical of the way anything Musk-owned is going to treat private data. I think he wouldn't hesitate to dig into it if it were to his benefit, even if there was an agreement against it. For that reason alone it makes Cursor look worse to me.
There's no way I will continue to use Cursor if it's Elon-owned. His actions at DOGE literally caused the deaths of thousands. We should all boycott.
hundreds of thousands
Musk's completely pointless destruction of USAID will kill millions of people.
Yep, and the majority of them are children. It’s so horrific that it deserves to be brought up any time Musk is mentioned for the rest of his life.
Citation needed.
https://www.thelancet.com/journals/lancet/article/PIIS0140-6...
Thousands is a conservative under count according to this.
>USAID
it's not Americas job to take care of the rest of the world.
Two things:
A) With great power comes great responsibility + making the world better through eradication of diseases, heightened food and economic security, improvements to the stability of nations, access to shared research and information, and increased personal agency makes it all better for Americans too.
B) Given you probably find (A) uncompelling, you should familiarize yourself with the concept of "soft power" and try to learn a bit about how relatively inexpensive it's been for the US to have an outsized influence around the world over the past many decades (outside of stupid, unnecessary wars, of course) and the benefits that's yielded for our economic and geopolitical interests across the board. If you'd prefer to not though, don't worry -- provided you're around for a couple more decades, you're going to get a crash course in the alternative.
That must be why Trump spent over $50B bombing Iran and agreed to pay them several hundred billion to go back to the status quo.
This comment is fundamentally not understanding USAID's role. You are not equipped for a discussion on this issue.
I'm Israeli. Trump worked with Netanyahu in trying to "convince" the people of Gaza to leave, so Trump could turn Gaza into a vacation resort.
Musk of course used twitter so Trump could get elected.
do also note that twitter is now the only place not censored by Israel.
How does Israel have the power to censor everything but Twitter?
>His actions at DOGE literally caused the deaths of thousands.
you mean that part where the US spends billions of dollars helping other countries and they hate us for our troubles?
pretty sure they hate us for other reasons and also reducing how much AIDS and ebola is in the world is good, this seems obvious to me
> The real question is whether SpaceX ownership improves Cursor or drives its users away.
I can't speak for anyone else but I wont be renewing my sub. Funding anything Musk related isn't exactly high up on my list of desires, and theres ample alternatives out there.
Yeah, he’s so damn successful at building companies that elevate civilization. Wtf would anyone want to go with that.
Sure, lets ignore everything else. Keep that head buried deep in the sand.
My sense is that enterprises are extremely cautious. They like everything that is already common and hr friendly. They abhor anything that might be seen as divisive and controversial. That's why they're currently going with Anthropic and not Openai or Xai or anything Chinese. It's the smaller actors that are using everything but Anthropic. Anthropic got that safe enterprise bland vibe. The only pr trouble Anthropic is in is with saying no to the military, which just makes them even more enterprise safe. Meanwhile Sam Altman and Elon are out there freaking out the enterprises almost every day it seems like.
Fable got halted by US make it a moot point but none enterprise is happy about the forced (and unannounced) 30-day data rentention
To be fair, the retention is only for Fable. I agree though, I don't even use Claude at work, but I noped out of touching Fable when I read that.
i’m not interested in supporting a trillionaire with my money. i’ll be moving to vscode since the cost of switching is basically zero.
I think it'd be an enormous endorsement of Cursor/xAI and proof of improvement if SpaceX started using it to code the mission critical software running on Falcon 9. Which other AI company can say their models powered a rocket launch?
(mostly /s but I know I'd give it another look if it was that good)
> Which other AI company can say their models powered a rocket launch?
Honestly, probably all of them. I imagine those coders are using all the tools they have available and are using Claude and ChatGPT as well as internal tools.
I just uninstalled it FWIW.
I can't imagine that any of the cautious companies or ones with their ear to the ground are going to want to cozy up to cursor with this acquisition; I'd suspect we'd see some exodus as well given the relationship to Musk.
IDK if you understand how much $60B is.
Tesla was popular before Musk's ketamine fueled schizo-fascist phase. People will probably boycott Cursor as soon as he buys it...
Funny that “GitHub should really support stacked diffs” led the Graphite team to a space colonization co.
2020: Leave Meta and start a company.
2020–2021: Spend ~18 months looking for PMF.
2021–2025: Build Graphite around stacked PRs, code review, and merge queues.
2025: Get acquired by Cursor because AI makes code review the bottleneck.
2026: Cursor gets acquired by SpaceX because Elon.
Not a startup arc I would have predicted from `gt stack submit`.
Ah, it's somewhat like how Jeri Ryan's divorce caused Obama to get elected.
Source: https://www.reddit.com/r/startrek/comments/187ghq/star_trek_...
I've tried most of the tools out there but have used cursor most consistently. Sure, some of the UI quirks get in the way sometimes, but I've found its auto complete predictions to be unparalleled. More importantly, these days I mainly use its Ask mode, Plan mode, and Agent mode. I like that I can use Opus via subscription pricing without Claude Code's wild and buggy harness. And I find cursor's plan mode to perform better than Claude's, but that may just be my personal preferences. I know cursor stopped being the cool thing a few months back, but I genuinely feel most effective with it!
But I'll stop using it now, for the same reason I wouldn't buy a Tesla, or support that maniac in any other way. And I'm sad about that :(
I used Cursor for a couple of months like a year ago, and I found it to be good, but I preferred the CLI experience and switched to Claude Code. "Claude Code's wild and buggy harness": not sure when you tried it, but I've been using it daily for the last couple of months, and okay there were some quirks here and there, but its stable and very usable.
Zed is pretty good these days. With a claude subscription, you get close to the same experience as Cursor, but with an editor that's much snappier.
You can’t use a Claude subscription in Zed. Codex and Copilot are supported
/edit: occurred to me that OP meant Claude via ACP, not directly
To be clear, you can totally use a claude code subscription in Zed. You just risk a chance that they're ban you and your subscription for it. Or you have to pay API pricing. Who's got money for that though, shit.
Signed up for Zed upon seeing this, signed in with GitHub, it failed instantly. Emailed support, they asked me to prove my ID because I was flagged as a bot. I immediately revoked access.
Fairly certain the support reply was from a bot, too. Meh.
$60b is crazy.
Cursor is an extension for VS Code, a harness and a bunch of prompts.
They have their own model (Composer 2) which is based Kimi K2.5, but I don't think SpaceX would be interested in it.
If they need a harness for grok, they could fork PI.
What are they after for here? Customer base? Talent?
Cursor has $4 billion annual revenue rate so $60b is 15 years of future cashflows.
That's not crazy because if past predicts the future, that revenue will grow quickly. At $8 billion/year it's just 7.5 years, which is a reasonable investment.
This is a linear regression relying on a couple years of data to predict 15 years in the future and I don't believe that the valuation is made on this basis.
It may be that spaceX is buying an operation that would realistically take 5 months and 100 million to copy in-house for 60B because the worry is that waiting 5 months might cost that much in some sort of lost opportunity. It also might be that in any negotiation SpaceX is viewed as incredibly cash-rich and so anything can be sold to them for inflated prices.
I really don't understand these companies valuations it seems like boardrooms everywhere are in a constant state of panic that they'll lose it all if they aren't growing a breakneck pace constantly.
Not annual, annualized. Let's see if people stick to it knowing it belongs to Musk now.
Many made the same prediction about Twitter, and it seems to be more or less the same or higher activity than before, and Bluesky is continuing to rapidly decline (-5% DAUs per quarter).
I wouldn't bet on Musk, but I DEFINITELY wouldn't bet against him. Anyone betting against him over the last 10 years has been viciously smoked (many short-selling hedge funds got wiped out completely).
> I wouldn't bet on Musk, but I DEFINITELY wouldn't bet against him. Anyone betting against him over the last 10 years has been viciously smoked (many short-selling hedge funds got wiped out completely).
In terms of the stock market, definitely. Honestly though, all those people who said self-driving wouldn't be solved by now, that Tesla didn't have a great moat and that the Boring Company was profoundly stupid were in fact correct.
> Many made the same prediction about Twitter, and it seems to be more or less the same or higher activity than before
Where are you getting that? There's not a single piece of data out there that will tell you Twitter has increased in users. Not only have they visibly dropped in users, but it's becoming increasingly clear the site is Astroturfed with bots.
The choice is not Twitter or Bluesky. Most people moved on to TikTok. I don't know a single person who uses Twitter.
Tiktok... what are you talking about? Tiktok is not a text news/thread medium
Why would that be germane to the discussion? Are you under the impression people are hardwired to use specific kinds of social media?
The thread is about Twitter supposedly *growing* in size recently, and its only evidence was Bluesky's userbase. That's just silly. I know half a dozen people who left Twitter for TikTok. They don't care about the slight differences in these platforms, they are all algorithmically driven platforms where people comment under text and video. They care more about what audience is cultivated and what is forced on them.
*jams head in sand angrily
> Many made the same prediction about Twitter, and it seems to be more or less the same or higher activity than before, and Bluesky is continuing to rapidly decline
Not at all correct, they are plummeting.
"In June 2025...X year-over-year growth declined by 15.2%." https://techcrunch.com/2025/07/07/threads-is-nearing-xs-dail...
> daily mobile users down 15% year-over-year
So just to be clear, down 15% YoY is not "more or less the same or higher activity"
the platform is overwhelmingly bots, so those "users" are likely in a server farm somewhere
https://techcrunch.com/2025/07/10/as-x-loses-its-ceo-daily-u...
Also, they reported a 60% drop in advertising revenue from 2021-2026... yikes
Twitter did lose significant revenues. It came ahead on profit because it aggressively cut costs which is an interesting thing to do for Musk given SpaceX. Maybe he learned that profitability is over-rated?
Yup, more activity... Elon never solved the bot problem like he promised, and it's far worse now.
I looked the other day, the feed got the same 4 postings reposted 3-5 times. A dozen posts talking about people getting banned, doxed and fired for talking about you know what.
I found 1-2 sort of interesting posts but not interesting enough to make it worth it.
It was fun for a while blogging with grok back when it was free.
Kind of interesting that engagement is zero on all of my postings. Not that I care for it but it shows quality isn't measured by users. The system prefer people read about users getting banned and that same video again and again. lol
With rss I just sort by date and eliminate duplicates. Not very hard and dramatically more interesting.
> Cursor has $4 billion annual revenue rate so $60b is 15 years of future cashflows.
This assumes that Cursor's annual revenue will be the same or higher for over a decade. It's not really like they don't have competitors
That revenue number is almost meaningless, since they give out tokens at a loss. Especially with Composer 2.5 tokens are sold at a steep loss. They could certainly grow to $8 billion/year, with this negative revenue / heavily subsidized subs, but what will happen if Cursor decide to be profitable, or maybe to even just break even?
If past predicts the future people will drop it once it is in Musks hands. And as a token reseller that revenue is not that impressive.
Is that profits or revenue? :-)
Hasn't cursor's revenue declined? I think they already peaked.
revenue ≠ future cashflows
> What are they after for here? Customer base? Talent?
users
And IP
It's a crazy number especially since Cursor feels kinda dead. Few thoughts from the other side:
- xAI needs the coding related data to compete with Claude Code and Codex
- Recent progress with Composer 2.5 seems promising given the size
- The may get a comeback on the smaller than Enterprise battle field now that the other two got so expensive
- The way that Elon set up this entire process was quite genius. They locked in this option before, and now after the gains through the IPO, it feels almost like a discount, lol
Compose 2.5 is the default model in Grok Build. And it's quite incredible. It's comparable to Opus 4.7 but faster and incredibly cheaper .
I wonder what happens to fireworks ai, who provided the infra to train and serve composer 2, cursor was their largest customer, and they're probably loosing it.
I could see it being a talent / first mover acquisition. I’m bullish on harnesses, but I think there’s still a very long road to get to something that is stable and relatively optimal - harness user experience is pretty trash tier right now imo.
My hot take is that it will probably be like the OS landscape:
It’s the coding, prompting, training data aligned w harness and ultimately usage of the user base.
Well. $60b in bloated stock is probably much better than the $10b cash penalty for not going through with the deal.
... you could just sell $10b in the bloated stock, and still have $50b of bloated stock to sell
shocked that an engineer would not mention user numbers or revenue growth in their analysis
Usage data.
Exactly that is the current value, although not sure if its worth 60B
You clearly have not used Cursor lately. It’s substantially more than all that. It’s not a VSCode extension anymore.
[flagged]
[dead]
Here's a shower thought. BTC essential is worth $70k solely through the power of memes. Can TSLA and SPCX remain overvalued (relative to the revenue of their respective underlying assets) forever through the power of memes?
Intuitively, it seems to me that there must necessarily be some kind of upper limit, but I'm not convincing myself. These speculative assets are only attractive as long as the price keeps inflating. But that can only happen if there is more and more demand. So it's basically a bet that there is an average amount of retail investors (I assume it's mostly retail investors but I could be wrong) that consistently put a percentage of their income into these speculative assets. Can this be maintained forever?
The saying "The market can remain irrational longer than you can remain solvent" exists for a reason.
In short, the answer to your literal question is "no" because nothing remains forever in this world. The practical answer is "yes" because the TSLA stock has been irrational for years already and it shows no sign of stopping.
Or the way I would think about it is (I think this was a Krushchev quote): You don't convince people there is no imaginary river, you convince people that you built an imaginary bridge across the imaginary river.
So you are never going to convince people that Tesla won't make a fortune on humanoid robots or SpaceX won't colonize Mars. The fever will break when people decide flying robots (or whatever) are a bigger business then humanoid robots or some other company than Space X will colonize the ocean before Mars. You aren't going to convince people to price these things reasonably but eventually the heat will wear off.
That saying assumes a rational market, and while there is some evidence that the average human behavior trends towards the rational there are plenty of evidence against individual behavior being rational (see homo economicus).
As more and more wealth get distrubuted to fewer and fewer hands, and as fewer and fewer extremely rich individuals control more and more of the market, My gut feeling is that if the market was ever rational (which btw. I am not entirely convinced of) that very much no longer holds true.
In the short term it’s a voting machine, in the long term it’s a weighing machine (Ben Graham).
The current episode of irrational exuberance will pass, as others have.
The before ones have had outside influence which forces the irrational episodes to pass. A spectacular collapse is usually fallowed by mass protests and labor actions, strikes etc. which forces the market to change its behavior. Most of the time these will result in a change in government policy which imposes regulation onto the market, and only then does the market go back to being rational.
> exists for a reason
The reason is that people keep making the same mistake, because people are not very good at assessing high-risk, high-reward projects.
Only a real deadline/delivery failure can wake people up, and only if they haven't pivoted their dream to something else, and only if you don't have other people knowing it's a scam, but willing to prop up the stock price because they are highly invested.
Scams, like cancer, are real; they survive, grow and defend themselves using the same mechanisms (laws, advisors, promotors) as ordinary investments/tissue, until they kill the patient -- so the best scams target the largest unkillable patients and enlist the broadest and deepest range of self-interested insiders as their defense.
It's beautiful, if you really think about it, as a tragic example of the worst of capitalism.
The thing is fundamentals really don't matter. TSLA and SPCX aren't paying dividends so there's no real performance they have to hit, no one is going to miss a dividend payment and dump the stock. The Elong vibes can carry it as long as people keep smoking what he's selling
The real question is, when does that run out of steam? When do we wake up to the charade that has built up around us? That's a much bigger thing than just Elon and his businesses. Like someone else said, when the next crisis/downturn/depression hits the house of cards will fall. Unfortunately it will hit all of us not just people in the meme stocks
I'm extremely cynical about the way the US government would react to any sort of financial crisis. I do not believe that they would not completely cave and bail out the AI companies and the monopolists if there is a downturn. And it's not that I don't trust the Republicans specifically. The whole political sphere is completely convinced that AI is a national security prerogative, and the cynical political atmosphere will equate national security with investor protection.
Let me append the saying a bit: The US government can remain irrational longer than you can stay solvent.
*The US government can remain irrational so you can stay solvent.
I’m pretty sure the market agrees with you.
Unfortunately.
If you exclude dividend paying stocks, then the entire stock market starts to look like a giant pyramid scheme casino.
> If you exclude dividend paying stocks, then the entire stock market starts to look like a giant pyramid scheme casino.
Stocks can start paying dividends in the future: MSFT did not in the past, and does now. AAPL did, stopped in the 1990s, and started doing so again.
You should be indifferent to the company's dividend scheme since it's the underlying business activity that drives total returns, and not its distribution policy. There is all sorts of magical thinking when it comes to dividends:
* https://canadiancouchpotato.com/2011/01/18/debunking-dividen...
* https://pwlcapital.com/the-irrelevance-of-dividends-still-a-...
A pyramid scheme can run out of people to keep it going: the stock market is in a sense a 'savings scheme' for future consumption. Younger people work and turn their cash into savings, older people take their savings and turn it back into cash: as long as young people need to think about the future, and older people / retirees need to pay bills, there's a mechanism to maintain this cycle.
> A pyramid scheme can run out of people to keep it going
And then you describe how the secondary stock market requires 'fresh blood' to whom to sell stock to cash-out.
It's precisely a legalised pyramid scheme that always needs someone to come in at the bottom hold the bag to let someone else cash-out. In turn they need someone to come in 30 years later. That's exactly how a pyramid scheme works.
> In turn they need someone to come in 30 years later. That's exactly how a pyramid scheme works.
The entire economy is a pyramid scheme: the expenses of some people (shelter, food, clothing, entertainment, etc) are the income of other people (landlords/mortgages/property taxes, farmers/grocers, etc). It's why, during economic downturns, personal virtues (saving) can become vices from macroeconomic perspective: if everyone is saving, no one is spending, and so producers/suppliers lose their income (and generally start laying people off, which causes more saving / less spending).
At any given point in time, if no one spends, then no one has income.
This was the 'innovation' of Keynes in the 1930s: use government spending to 'induce' demand to get the cycle going again:
* https://archive.nytimes.com/krugman.blogs.nytimes.com/2015/0...
For a stocks point of view: if no one is currently saving, then those that need income will lose it. At any given point there are folks who need to save/buy and those that need to spend/sell.
Exchange of goods with money isn’t a pyramid scheme imo. Financial system is a pyramid scheme, economy isn’t
A typical pyramid/Ponzi scheme falls apart when (new) people stop putting money into the system.
What happens to the economy when people stop putting into the system (by saving, and not spending on goods/services)?
> it's the underlying business activity that drives total returns, and not its distribution policy
That's exactly the question, though, since a lot of stocks seem priced disproportionately to their business activities.
> A Keynesian beauty contest is a metaphorical beauty contest in which judges are rewarded for selecting the most popular choices among all judges, rather than those they may personally find the most attractive. This idea is often applied in financial markets, whereby investors could profit more by buying whichever stocks they think other investors will buy, rather than the stocks that have fundamentally the best value, because when other people buy a stock, they bid up the price, allowing an earlier investor to cash out with a profit, regardless of whether the price increases are supported by its fundamentals and theoretical arguments.
* https://en.wikipedia.org/wiki/Keynesian_beauty_contest
Plenty of folks may think these companies are garbage but are 'playing along' because it's not necessarily what they themselves think that is important, but what others think. You can make money in a bubble, even when it eventually pops. What we're seeing now is hardly new, either:
* https://en.wikipedia.org/wiki/Technological_Revolutions_and_...
This is why I stick with index funds, as I don't really can't be bothered playing the game:
* https://en.wikipedia.org/wiki/A_Random_Walk_Down_Wall_Street
I generally check my portfolio once a year, in January, when I top things up when new contribution room becomes available with the new year. It's 'fun' to follow along with the gyrations and drama as things happen, but I don't sleep over it. If you're reasonably diversified you can generally weather storms and come out okay on the other side:
* https://awealthofcommonsense.com/2014/02/worlds-worst-market...
* https://www.forbes.com/sites/advisor/2010/09/13/its-not-real...
I don't have better ideas than you to do anything except "buy indexes".
Everyone is happy enough to give Elon (and others) more and more leverage to buy politically strategic companies (this is not that, this is probably just an ego buy for him, something to kill time because he can).
I was worried about him selling out (from an overall market and even index perspective when they were going to bend rules), but it looks like largely the whole situation is predicated on the idea that he can't or won't sell. I don't know how exposed the market is but it doesn't feel good.
Well, if we taxed the top percentages of income/wealth more, and actually enforced anti-monopoly rules, perhaps we wouldn't be in this situation.
> A Keynesian beauty contest
No shit. That's why, even if it's an exaggeration to call the entire stock market a pyramid scheme, you can't justify the claim that it's entirely "underlying business activity that drives total returns". That's the real question (from which dividends are, yes, a distraction).
> […] you can't justify the claim that it's entirely "underlying business activity that drives total returns".
The S&P 500 index tracks earnings per share (EPS) fairly closely over the decades:
* https://www.macrotrends.net/1324/s-p-500-earnings-history
A lot of folks think the top ten stocks in the S&P 500 making up ~40% of the capitalization is bonkers, but they also make up ~40% of the net income share:
* https://en.macromicro.me/collections/34/us-stock-relative/14...
So from an earnings/income perspective, there appears to be a link between the two.
Perhaps worth noting that the US markets seem to (only?) outperform when tech is outperforming, with other US non-tech sectors basically performing the same as out countries' non-tech sectors:
* https://ofdollarsanddata.com/do-you-need-to-own-internationa...
What do you think you're arguing against?
No, there are a huge number of companies that are valued reasonably for their revenue / profitability / growth.
There's basically two stock markets: things valued on fundamentals and things valued on vibes.
And I don't think there's ever going to be a unified theory of value that can span both, because the former is quantitative and the latter is psychology.
True, and it's a common question that comes up: What's the point of owning a stock if it doesn't pay a dividend? How do you participate in the company's performance?
There are other ways for performance to translate to the investors directly. For example, if the company sells itself then all the shareholders will get that payout. Stock buybacks are a thing. And, as other commenters here have said, eventually the company may start paying a dividend
But, you're not entirely off-base in that you're just buying into the vibes of a company. It's just that most of the time (much of the time?) those vibes have been rooted in some semblance of rationality, that there was something of value behind the shares you're buying. That is definitely not universally true anymore
There's taxation reasons. Stock buybacks mean that investors get the usually lower capital gains tax rates instead of dividend income tax rates. On top of this, it also juices earnings per share.
>How do you participate in the company's performance?
Most stocks give voting power even if they don't pay dividends. Notably, SpaceX's don't.
Aren't buybacks the new dividends?
It’s only a pyramid scheme for Tesla, SpaceX or X formally known as Twitter see a pattern there?
Gamestop is not an Elon venture. There's plenty of irrational stock valuations out there, most of them just come back to Earth on a reasonable timeframe
Would you say the same thing about Bitcoin? Will that house of cards all fall in the next crisis/downturn/depression?
Hasn’t it crashed in every economic downturn and financial panic so far?
Not really, no.
A decade ago it was under $1000 and has never been that low since. It's peak price is only about 2x the current price.
Isn’t a 50% reduction pretty bad?
And being higher over 10 years has little to do with it if acts counter cyclical to stocks and other assets.
> And being higher over 10 years has little to do with it if acts counter cyclical to stocks and other assets.
BTC has been called many things at many different times. It was originally a payment system:
> A purely peer-to-peer version of electronic cash would allow online payments to be sent directly from one party to another without going through a financial institution.
* https://bitcoin.org/bitcoin.pdf
And it can still be used for that, however the transaction throughput is tiny, and so it became a store of value in essence: but it's kind of hard to be that when the value swings up and won so much. While "fiat" currency inflation is annoying, it is, generally, fairly predictable in most cases (<4%) and so you can plan ahead with regards to future value and purchase. The same is hardly true of BTC.
Is it bad? Isn't 60x in 10 years still kind of ok? The things I invested in ten years ago didn't return 60x since.
Again, the original argument was that it was like gold, and provided a hedge against equities.
As the sibling says, the original value proposition of Bitcoin was that it would supplant Paypal and wire transfers. The store of value narrative was a post hoc rationalization by stakeholders when it became clear that it was technically unsuited to serve Internet-wide throughputs. Not only that. I sat in during the meetings around the time of the segwit transition; I think they were debating making the blocks bigger or something. It was amazing to see first-hand how such a relatively minor and technically necessary change could generate so much friction. That was when I realized Bitcoin was never going to evolve in a useful direction, and sevenish years on I see that I was correct, as the block structure is pretty much the same as the last time I saw it.
Yeah, now both the first and second rationalization have fallen.
Who argued that? In 2008 Bitcoin was proposed as a peer-to-peer digital cash system. The comparison to gold didn't really take off until years later.
I mean the original thing I replied to. The idea that it was a stable or anti-cyclical store of value emerged after it became clear the payments stuff was not working.
There is some misunderstanding of what a crash really is. It doesn't necessarily mean that things get written down to 0 or some arbitrary level because everything has a price at which someone will buy.
Even companies have some value after a crash and you could make a case that at some arbitrary point it was worth $x and since the crash didn't cause the company to crater to below $x it has not "crashed". Even companies filing for bankruptcy have some residual value above what they might have been founded on - it doesn't mean the company hasn't gone bankrupt.
Yeah, I certainly did not mean go to zero, or near that.
To be more specific, I have often seen people argue for including crypto in a portfolio based on the theory that if equities drop a lot (25, 30%?) crypto will hold or go up. People make the same argument for gold.
I personally think Bitcoin will eventually peter out, and that it's fairly risky as it has definitely been pushed by vibes and promotion and everyone wanting to get in on it
I mean that can apply to anything. There's nothing intrinsic about gold that makes it valuable other than it's rare, but there's plenty of things that are relatively rare that aren't valuable. Yes there's industrial uses of gold but that's not why we as a society and a species treat it as valuable
Maybe bitcoin is the new gold and will hold value forever, and as more serious people get into it, it will lose its volatility and be less subject to the vibes shifting and there being a run on the market. Maybe not
It is different from Elong stocks, though, because the myth of Satoshi aside, it's not exactly a cult of personality like his companies are
To me the difference is that BTC's continued value isn't predicated on the actions of one person, who has continually proven himself to be childish, mercurial, and dishonest.
Yes when will everyone wake up for the charade of having a monopoly on space launches? Of putting over 90% of all mass into space, of your closest competitor being the nation state of China, and they are years away from where you are right now. Ah yes, that charade, when will people learn am I right? Total genius.
An extreme majority of Space X's self-determined value proposition has nothing to do with space. Out of an overall TAM of $28.5 trillion, $26.5 trillion is based on AI. Personally I believe that number will actually approach zero, but it certainly even in my most optimistic view is a number with a B, not a T
The total addressable market even at Space X's own calculation for space launches is only $370 billion. And, supposedly as the only company that can launch things into space they are still losing money on that business. This is bankrupt-a-casino levels of incompetence
SpaceX are currently losing money on the "going to space" part of their business.
They're making money on telecoms, and may have just started making a profit on renting out the data centres they originally built for the AI that it turned out hardly anyone wanted to actually use.
Even if one assumes this is true, it still wouldn't justify the valuation.
The demand just isn't there, which is one of the reasons SpaceX is pretending they will generate more demand with their cosmic wizard spires.
Michael Burry, a hedge fund investor featured in the book “The Big Short” for his predictions on the 2008 financial crisis, said in a Substack discussion last month that any increase in SpaceX’s stock after its I.P.O. would “be on hype and technicals.”
Here, “technicals” means technical analysis signals rather than the company’s business fundamentals. In other words, Burry is saying the stock could rise because of chart-based trading, momentum, and market behavior—not because investors think SpaceX is truly worth that much based on revenue, profits, or other fundamentals.
How long can the hype be maintained? TSLA is still maintaining its hype, judging by its P/E ratio.
Technical Analysis is Astrology, and Burry has predicted 17 of last the last 2 stock market crashes.
> Astrology
Exactly, but there are people out there who buy stocks based on technical analysis. It can have an effect if enough people act on it's "signals".
The hype already starts with the SpaceX SEC filing. According to it, its addressable market is $28.5 trillion, of which $26.5 trillion are AI. This means that every human being who owns a computer on this planet (1.75 billion) would need to spend (on average) over $15.000 on xAI products.
>> that every human being who owns a computer on this planet (1.75 billion) would need to spend over $15.000 on xAI products.
Current US national debt is approximately $39.22 trillion. As we achieve a zombie movie, level of collective madness, lets take all this into the last degree.
Nationalize SpaceX, and with this bright obvious future described on the prospect lets pay US debt :-) Pay US retirement benefits pensions in token credits, pensioners can resell, make an options market for tokens. I am sure Robinhood will open you a margin account for that?
Lets open a futures markets for food goods grown up on SpaceX Asteroids, as they will have free solar energy. They can grow them three times faster than on Earth...
To quote Ron Baron yesterday on CNBC, we are all going to earn hundreds of billions....
> would need to spend over $15.000 on xAI products.
For perspective - that’s 12.5 years of Tesla FSD subscriptions. I think there are probably about that much cars out there.
Yes indeed.
While this would be conceivable if some future AI gets good enough to actually replace 100% of global paid labour currently done by using a computer, the reference class I have here is that Wikipedia is definitely not valued at [number of people online] * [peak cost of Encyclopaedia Britannica].
Economic displacement on that scale breaks the valuation.
Technicals are the star signs of stock market.
Funny, the European Central Bank just made life even harder, by raising interest rates purely on that, and the FED is preparing to do the same...
Fed is not raising rates because of technicals. It’s because of economic numbers. The two aren’t the same.
At the essence they are the same, if you think a bit about it.
they literally are not.
Looking forward to the explanation...
It’s not our responsibility to provide counter arguments. You expressed that the two fundamentally different things are the same “in essence”. It’s your responsibility to define the two and explain how they are the same first.
You, and the majority, are confidently wrong. Its only your mistaken confidence, that make you believe you dont need to provide a counter argument.
...and here I was hoping for yours.
Tesla definitely is floating down slowly worldwide with hype when it finally crashes just don’t be left holding the bag. 2026 will be another year downward.
I guess now a lot of Elon fans change from Tesla to Spacex.
[flagged]
The market will stay irrational for longer than Michael Burry will stay correct.
Notably, it looks like the meme power is gradually being redirected from BTC and cryptocurrency in general towards AI and SpaceX in general. Now that people have found a means of consuming vast amounts of computing power that occasionally emits useful output, rather than just a hash and a colossal waste.
> Here's a shower thought. BTC essential is worth $70k solely through the power of memes. Can TSLA and SPCX remain overvalued (relative to the revenue of their respective underlying assets) forever through the power of memes?
The "value" of something can be a bit of a meta-game:
> A Keynesian beauty contest is a metaphorical beauty contest in which judges are rewarded for selecting the most popular choices among all judges, rather than those they may personally find the most attractive. This idea is often applied in financial markets, whereby investors could profit more by buying whichever stocks they think other investors will buy, rather than the stocks that have fundamentally the best value, because when other people buy a stock, they bid up the price, allowing an earlier investor to cash out with a profit, regardless of whether the price increases are supported by its fundamentals and theoretical arguments.
* https://en.wikipedia.org/wiki/Keynesian_beauty_contest
Plenty of folks may think these companies are garbage but are 'playing along' because it's not necessarily what they themselves think that is important, but what others think.
This idea was put forward by Keynes in his General Thoery publish in 1936, so human nature has hardly changed since then.
While gold has industrial uses, isn't most of its value based on the fact that people like it and believe that it has value?
I'm not sure if this means gold isn't powered by memes or whether it's just one of the most long lived memes of all time. Aside from other nice properties like lasting a long time, being pretty, and not requiring electricity to exist.
Certainly, gold is not a meme, nor are all the other rare earth metals. This principle applies most to the other elements found on the periodic table. In fact, they have become more valuable due to research and development in today’s modern world, as numerous useful discoveries are made daily.
Cryptocurrency, such as Bitcoin, is undoubtedly a meme. If given a choice, one would prefer to possess any of the rare metals, rare earths, or any other elements that can be obtained in sufficient quantities, whether on land or in the ocean. Bitcoin or cryptocurrency is beneficial only to insiders and not to the general public. In essence, they are akin to modern tulips with a cherry on top, and like Tesla in 2030, one should avoid being caught with the bag.
What you are describing applies to all forms of money. It has value because people believe other people will use it as money. If that belief drops, the value drops.
People comment on gold and Bitcoin, but don't realize the same principles apply to US dollars and bonds.
> the same principles apply to US dollars
Currencies are a little different, since they are required to pay taxes; and payment of taxes is enforced (to varying degrees) by state violence.
Hence if you believe you will be taxed ("death and taxes" being the only certainties in life, etc.) then the currency associated with that tax has value, in that it avoids imprisonment, etc.
I was once told bitcoin was a 'pet rock'. Thing is people pay a lot of money for rocks when they have no planned industrial activity for them. Diamonds, for instance.
I think you are spot on. The problem comes if SpaceX goes out of fashion, not its fundamentals.
[dead]
This is why I think Elon buying Twitter was one of the most strategic decisions he ever made: he quite literally bought the meme machine that upholds the valuations of his companies.
I still don't think it was strategic on his part, he did try to back out after all; my guess is that having bought it, he tried to maximise the value it produced.
And given the financial statements in the SpaceX IPO, to the extent X still has any value at all, it is almost all just influence of one kind or another, not actual money.
And the government that helps fund and regulate (or not) them.
At least as long as Musk is the CEO, perhaps. I don't think it's easy to find another charismatic figure to keep it going.
It's a different kind of hype than Nvidia has, which is showing very high and fast growing revenues (which may not continue, but they're real now). Jensen I think is not as critical to the AI hype as Elon is to his companies.
All these major tech companies eventually get leadership changes. Apple, Google, Amazon, have all done well because they're real companies and go beyond their original leadership. Tesla and SpaceX I think would surely go down the moment Musk is no longer in leadership.
you think Musk is charismatic?
https://www.natesilver.net/p/elon-musk-polls-popularity-nate...
majority view him unfavorably.
I think we've all seen by now that the majority doesn't much matter if you have a rabidly zealous minority.
suppose you've got me there
> majority view him unfavorably.
Indeed, but for general stock market purposes it's one dollar one vote, not one person one vote.
Possibly even worse. Short-sellers got burned on Tesla (and perhaps now also SpaceX), which may mean the marginal buyer and seller consists only of people who already buy into the hyper and by trading are sharing price info with each other rather than with a single person who doubts the man.
I sold my Tesla stocks a while back; the people who kept them don't care what people like me think.
Well I meant he has a big Twitter following and whatever he gets into many are willing to fund and so on. Like there's no way Tesla would be taken seriously for all its future ambitions and valued accordingly without Musk at the helm.
I just cancelled my cursor, which was the best for my workflow. Will need to find something else.
same just cancelled it and left Elon Musk as a reason
> Can this be maintained forever?
Obviously not. It’s all about timing your bail so you don’t get left holding the bag.
On what grounds do you believe the value of BTC is meme-driven? Another (and arguably more plausible) explanation is that the price reflects the vast amount of criminally-obtained wealth stored in it. It’s a far better store than burying cash in mountain caches.
All money is meme driven. Money is fundamentally a meme itself. Seeing lots of people who seem to be making some sort of implicit distinction between bitcoin and USD and so on, but they are no different. They serve the function they serve because of what people believe about them, like any other social, cultural, or economic abstraction. Bitcoin has the feature of a verifiable ledger, but its value and function are in our heads, just like the USD or GBP.
tax payment is kindof a meme i guess, but one with force behind it
BTC is a terrible place to put criminally-obtained wealth. Public ledger and all that makes it very difficult to hide anything.
That ship sailed long ago, before there were viable alternatives to BTC. There are plenty of ways to launder BTC (i.e. hiding in plain sight) and the existence of hoarded wealth in BTC isn't necessarily exclusive of other cryptocurrencies that are better fit for purpose.
> BTC essential is worth $70k solely through the power of memes
This is not true.
BTC is way more sane than SpaceX as can be seen by it's history so far.
SpaceX is successfully building reusable rockets. BTC can process few transactions every ten or so minutes at enormous price.
I think SpaceX is definitely overpriced but saying that BTC is more sane is completely delusional.
BTC has the layer 2 lighting network that can process more transactions than Visa and MC combined with very cheap fees. I guess HN knowledge of Bitcoin does not go further than 2015
"Can" != "does". The current actual throughput of the Lightning Network is estimated at around 300 TPS: https://www.spark.money/tools/lightning-vs-visa
Also, everything on level 2 needs to be ultimately balanced on level 1; BTC doesn't have enough throughput to balance all the banks in the world, and lo Lightning appears to be consolidating into fewer nodes.
Also also, everything level 2 (including Lightning) necessarily takes away at least one of the selling points used for BTC, and replaces them with something functionally equivalent to 50% of a bank but worse.
It literally says this in the website your linked
> Current actual throughput is significantly lower, estimated at around 300 TPS, because the network is still growing. However, Lightning's architecture scales horizontally: as more nodes and channels are added, capacity increases without any protocol changes. The network's total value locked has grown past $500 million, with over 15,000 active nodes and 50,000+ payment channels.
The layer 2 has been predicted by one of the first adopter in 2010 so it was always the plan
https://bitcointalk.org/index.php?topic=2500.msg34211#msg342... It can scale without any changes and if there is demand it will surpass Visa and MC
SpaceX the company has done and is doing amazing feats of technology.
SpaceX the stock is only in a very small part about rockets and space. In their filing they talk about their total addressable market being $28.5 trillion, of which $26.5 trillion are AI.
That means that they believe the SpaceX stock is 93% about AI.
It's open to argument whether xAI or BTC is crazier - I just wanted to say that when talking about SpaceX stock, it's not really about rockets.
dont think the value of btc comes from its utility anymore
stables are much better for transactions anyway
neither of them are sane. BTC is useless, unless your trying to buy child porn, buy illicit drugs on the internet, or someone who bought it before the value exploded. eventually, the world will come around and it will go to zero, if quantum doesnt kill it first. im looking forward to that day.
>BTC is useless, unless your trying to buy child porn, buy illicit drugs on the internet
The market has long shifted to "buy Litecoin with cash, swap it to Monero" for these kinds of activities.
i know exactly zero people using any of these markets, and 99% of the other people here can probably say the same.
USD is useless unless you want to buy something with it. You can't eat it, it has little fuel value...
...it is also propped up in value by a frankly insane demand for it from other countries...on its fundamentals it would be worth much less
what is your point?
[flagged]
BTC's use is being bought and sold. So it's not useless. If I offer you one BTC for $30k, will you buy it?
If I offer you a tulip for $25,000 would you buy it?
Sure, if there were a stable enough market in which I can fetch twice that amount for that tulip, I'd buy it in a heartbeat. Wouldn't you?
The point being: BTC is a an abstract good, of no practical use except that of being transacted. Has whatever value the people are willing to pay for it, and has had a value in the tens of thousands for long enough that buying one with the intent of keeping it for a while is not such a stupid idea. I don't currently own any but there is a price at which I would buy one, and that price is many thousands of dollars... For an alphanumeric code in a distributed ledger.
Being transacted is the least usable thing for BTC. He you tried to transact on it? Its so slow its funny people claim it a use case for it... I know of only one use case, and its to anonymously bribe politicians who have their own coins.
I didn't say it's used to buy things. I said it's bought and sold.
If I think the resale price will stay at $30k long enough for me to flip it, then obviously yes?
sure, because you'd be offering me an idiotic arbitrage. your sentence doesnt prove it's useful.
It is not useless, it is incredibly arrogant to think it is useless at this point. You can do a 10 minute research and see how useful it is
True, it's very useful for scammers, grifters, international terrorists and authoritarian governments to funnel monetary value around without having to rely on the traditional financial system where they may be blocked and their money confiscated.
It costs 150 usd to send 5k usd between two countries I live in.
It costs the same amount of money to literally go there by plane and bring it as cash. This is not fair obviously. So traditional finance is a scam itself.
It takes 10 seconds and no fees to make the same transfer via blockchain.
I am pretty sure you know basically nothing about those crimes or the people that do crimes or how they actually transfer money. Just doing casual newspaper intellectualism while talking about things you never interacted with
what percentage of BTC transactions do you think are being used for international cash transfer in 2026 ?
> funnel monetary value around without having to rely on the traditional financial system where they may be blocked and their money confiscated.
Yes. Unfortunately the traditional financial system is governed by a country which has been behaving increasingly erratically, threatening its long-term allies with invasions and committing obvious war crimes. This is not a "good guys vs bad guys" scenario.
As long as there is someone around who is very good at keeping the price inflated (and that in turn also because he did actually deliver extraordinary things, it's not just smoke and mirrors).
On the other hand, the fact that BTC has absolutely no intrinsic value can be an advantage over a real company, as it makes it more insulated from reality. Supply chain shock? No problem. Competition? Same. New technologies, political change? Neither.
USD has no intrinsic value either
At least that one is backed by a military.
The current market cap of SPCX and TSLA combined (~$3.8T) is about 3 times the total value of all BTC (~$1.3T).
SPCX only floated a small amount of shares, and made the stock exchanges compete so that it would get to rig the system in a way. If funds have to buy SPCX + small share amount, it's going up. This is the reason stock buyback used to be illegal. It's purely market manipulation of share count, not market based on actual value.
I saw that one wealthy individual had purchased $1 billion of SpaceX at the IPO. Does that count as a retail investor?
> I saw that one wealthy individual had purchased $1 billion of SpaceX at the IPO. Does that count as a retail investor?
There were two individuals who each bought $1B: Ron Baron and Gina Rinehart.
While they are individuals, they executed these billion-dollar investments through their massive corporate entities and investment firms, rather than personal brokerage accounts.
A retain investor is an individual, non-professional investor who buys and sells securities through brokerages using personal funds.
At least some of TSLA and SPCX value is derived from Musk's ability to purchase politicians to ensure the tax dollars keep flowing to them. Essentially, these ticker symbols are backed by the US government's ability to force us to give Musk's companies our money.
>>Intuitively, it seems to me that there must necessarily be some kind of upper limit
I mentioned this in other thread(https://news.ycombinator.com/item?id=48514481), we are at runaway intelligence already.
Mostly because we are looping AI to fix problems, and then the same data is used to improve AI. There is no upper limit to this.
Taken to its logical conclusion, this process needs a hardware scale that might even look laughably huge at this point. Its fairly obvious space is going to play a big role in the coming times.
I could be wrong, and I humbly accept it when Im proven wrong. But it does feel like a lot of people in top places know we are going to need all the energy and resources space has to offer to run this runaway intelligence.
> we are at runaway intelligence already. Mostly because we are looping AI to fix problems, and then the same data is used to improve AI. There is no upper limit to this.
I'm an AI booster, but this really doesn't follow. There's absolutely no guarantee that the marginal improvement from this continues to hold.
Its not marginal improvement.
Validated data coming out of LLM(Itself generated from a recursive/loop process, used to incrementally arrive at solutions) being using to improve the very LLM is a very powerful loop. And there is no real upper limit to this, at least not in the near future.
Like most exponential processes, the start is slow, but it gets fast very rapidly.
This is also what Anthropic has said, the ones training these models today(i.e 2025 - 2027) will be impossible to catch up with, let alone beat.
So we are at a kind of runaway AI already today.
Economics meaning of marginal, not the qualitative one.
> there is no real upper limit to this
This is what you need to cite for your argument to hold together.
It seems like that during a bubble (things can never go down!) but the market eventually does correct, even if it can take years or even decades to do so, and usually overcorrects when it does.
It's down to the balance between buyers and sellers. If you've got more selling than buying it'll go down but Musk has been remarkably good at keeping the buyers there.
Don't forget the amount of shares released to the market. That impacts the price dramatically as can be seen be all the 'stock buyback' manipulation of stock prices.
Indeed. There's bit of a funny effect just now because the various index funds are having to buy into SpaceX so their funds match the index and apparently that's ~$14bn of forced buying at a time when there are not so many sellers.
Pump and dump only has to meme hard enough for the pump part.
If there is ever another depression, which seems highly probable at this point, the meme stocks will be the first ones down the toilet.
Everything is ultimately valued by memes.
money is just an idea that spreads (a meme).
Sure until retail are forced to sell in an correction. Wealth destruction can happen very fast
Bitcoin isn’t valuable solely through memes, it’s also deflationary by design.
Someone invents something that is digital, but can't be copied. Quite brilliant as it is the first digital asset that can store value without centralization and trust, based on market demand.
Someone on HN: "BTC is valueable solely through the power of memes".
Bitcoin doesn't "store value", it has the value that society assigns it, which is what the parent means when s/he says "BTC is valuable solely through the power of memes". It's not a fiat currency, nor does it have any intrinsic utilitarian value.
Age old argument: money also has no intrinsic utilitarian value.
And yes, btc does store value, it is doing that for me now. I stored some of my value in it and it has held better value than fiat.
> Age old argument: money also has no intrinsic utilitarian value.
I didn't claim money had intrinsic value, I claimed that money is fiat currency--it's backed by a lot of people with a lot of guns. BTC is not.
> And yes, btc does store value, it is doing that for me now. I stored some of my value in it and it has held better value than fiat.
Currencies (BTC included) don't "store value", they are assigned value by society esp via markets. Sometimes society decides one currency is more valuable than another, or that its change in value differs from that of another, but neither currency stores value.
You mean like gold?
No, gold is not a fiat currency and it also has significant utilitarian value.
The overwhelming majority of SpaceX holders are institutional. They had planned to allocate 30% to retail, but it ended up in the 20% range as a result of institutional demand. [1] No clue what's going on right now as their stock is going to the Moon. But in any case, I think the people that don't understand why it's doing well are mostly those who are unfamiliar with the space industry. SpaceX has already revolutionized space by dropping the cost to orbit by multiple orders of magnitude relative to the Space Shuttle, which it replaced. And it looks set to do that again with the Starship.
The main reason humanity hasn't meaningfully started expanding into space is because it used to cost $54,000 to get a liter of water into space. SpaceX brought that down to about $5000, and then further down to near $1400. That's a massive reduction in price, but you're still left with the problem that it costs $1400 to get a liter of water to space, which is why we still can't have nice things, yet.
Starship has the promise of bringing that down a couple of more orders of magnitude where the goal is to get it within the $10-$20 range. If they succeed, then you've just opened the doors to an entire new frontier of expansion and growth for humanity which is practically infinite. And right now there's no real reason to think that they won't succeed. And more importantly than this is that nobody seems to be able to compete on their level, or even remotely close. Their closest competitor is probably China who remains technologically well behind. And so SpaceX today is akin to being able to get a piece of some sort of super-ship making monopoly, just prior to the Age of Sail. The downside risk is basically zero since they're still making rapid progress - the only question is how rapid. And upside potential is basically infinite.
[1] - https://www.cnbc.com/2026/06/11/spacex-cuts-retail-ipo-alloc...
> The main reason humanity hasn't meaningfully started expanding into space is because it used to cost $54,000 to get a liter of water into space.
Am I naive in thinking that we haven't expanded into space because we don't need to? What's the benefit?
> What's the benefit?
For why we'd want to go at all: there's a lot of resources up there, and pollution is much less of a concern for factories made up there. Also some material processes may be much easier in zero-gee.
But that doesn't mean it's actually worth the effort.
Asteroid mining is the obvious one. That said as part of looking into SpaceX's business I learned that currently the TAM for launching other people's stuff into space is under $10 billion. SpaceX already owns most of that market. Their own focus on data centers in space IMO speaks for itself.
A hedge against civilization-destroying events. Meteors, global war, disease.
Space is not a hedge against these things. At best it's our final tomb.
There is no backup plan.
to be fair, in the pharoahs quest for immortality after death, if they could have put their pyramid burial tombs in space to avoid decay/ruin for longer they would have. I wouldn't be shocked to see the oligarchs paying for their own space burials inside very lavish satellites whose orbit won't decay for thousands of years so they can "live forever".
> we haven't expanded into space because we don't need to? What's the benefit?
Access to resources. A sense of adventure. Learning.
Which resources are cheaper to get from space? I’ve only ever seen hand waiving arguments, never any math.
That depends entirely on the price; things are very different when it costs $20k/kg to orbit (basically nothing is worth that) vs. $20/kg to orbit.
Some people are suggesting making high quality fibre optics and pharmaceuticals in LEO already with Falcon prices:
https://en.wikipedia.org/wiki/Space_Forge
https://en.wikipedia.org/wiki/Varda_Space_Industries
While I am *extremely* skeptical, I've heard people talk about space-based solar. I mention this only so nobody adds an "and also" for that.
Ditto anyone talking about lunar He-3. This is a bad idea, "why" is long enough to be worthy of a blog post.
But if you were committed to expansionism for whatever reason, then you'd want some way to get industrial quantities of steel from mines already in space, to make the factories themselves cheaper. Possibly megastructures like an orbital ring, too.
The hand waving is probably reasonable for a couple of reasons. The first is that you can't really math things out yet when costs are shifting by orders of magnitude on relatively short timeframes. What seems impossible in one year is a casual achievement in the next. If we go back to 2002 (when SpaceX was founded) and I told you that within 2 decades they'd have lowered the cost to get stuff to space by more than 97% relative to Space Shuttle, it'd have sounded somewhere between unreasonable and impossible, yet that is exactly what happened. And Starship is set to send prices plummeting yet again.
The second reason is that the notion of economics becomes kind of odd when you start introducing space. Space has practically infinite resources of every type imaginable. For instance one single asteroid, Psyche 16, is thought to have orders of magnitude more gold than has ever been extracted in human history. But I mean, what does this even mean economically? Obviously if you start bringing back hundreds of thousands of tons of gold, the price is going plummet. And even the viable possibility of this happening would probably send the price of gold plummeting. So... it's weird. We're in uncharted waters.
And then on top of this access to space will mean we will start expanding outward - colonies, vastly larger stations stations, and so on. And these expansions will develop their own parallel economies with their own needs/production/etc. This is all kind of like trying to predict the economics of the internet before the first cable had been laid.
Most people I see online freak out at the idea of normalizing living in an apartment/condo and slightly higher density. I don’t see them lining up to live in space :)
As far back as the 60s NASA had already laid out technical plans for how to get humans to Mars. The Moon was never meant to be the destination, but a brief stepping stone to greater things. So we can thank Nixon for setting humanity's progress back by half a century. The ISS and Space Shuttle both were driven largely by these initial plans.
But the current claustrophobic nature of the ISS was, again, not the destination. There was a grand scheme of in-orbit refueling, fuel depots, and much larger scale space architecture. Something of the ISS scale was intended as a short jumping-off point, a workers' tent analog. But our progress stalled out and the workers' tent is what we were left with.
The point I'm making here is that space in the era we're stepping into is very unlikely to look much like the one we're leaving. This is even more true as the price drops because private industry will be able to play a major role. There will be no Nixon to just cancel everything out of concerns for his political career.
Depends on the timescale, and what other tech is the prerequisite to this and what other things can be done with that tech.
Even with cheap rockets, Musk was talking about 6-digit-dollars for becoming a Martian, I don't see more than a few million people wanting to spend that much for that outcome.
If we get self-replicating robots or whatever, big space stations aren't that expensive, but also Earth is huge and you can extract minerals from seawater electrolytically and PV is already cheap, so perhaps we'd just get a load of seasteads instead: https://en.wikipedia.org/wiki/Seasteading
How far out would you say that is? Have you sketched out the net present value calculation for an investment in spacex today based on that?
> How far out would you say that is?
Really hard to say. Everyone who does space stuff talks with the same level of optimism as Musk but historically they deliver less.
But for what it's worth, I don't think Musk's going to reach Mars except in an urn, baring some surprise development in anti-aging medicine. Even though their progress is rapid by the standards of the industry, they are still slow in terms of natural human lifespan and his age.
> Have you sketched out the net present value calculation for an investment in spacex today based on that?
IMO, their market cap should be around $200bn at the moment. That's mostly Starlink, so if there's a Kessler cascade *or* if China makes a competitor that goes down to about $80-100bn
The long-term stuff about Mars or turning the moon into a factory for modular data centres is too far into the future to be worth considering.
That was my general impression. They did amazing work to totally change launch costs (with more on the way). But launch was a limited market. The low costs have and will grow it, and starlink will grow it, but not enough to justify this value.
In investing it's critical to think about things probabilistically, rather than trying to just guess the future. "Fooled by Randomness" is a great book on this. In a scenario like this it's especially critical, because in the case where Starship fully delivers and the next great frontier of humanity opens up, the upside potentially is practically infinite.
E.g. imagine in a bet you think there's a 99% chance of something being worth $1 and a 1% chance of it being worth $1 million. What's a fair price? It's a simple calculation - 0.99 * 1 + 0.01 * 1,000,000 = ~$10,000. That's you setting a price of $10k on something that you fully expect to be worth $1 99% of the time, and still coming out just fine.
So the actual debate should not be on 'guess the future' but rather on the odds of Starship delivering and estimating the impact this would have. And I think to many the answers are fairly easy - 'very nonzero' and 'very big.' It's akin to the NVidia/Apple stories, except in this case it's being somewhat priced into the market to begin with because it's somewhat more probable, and easier to foresee.
The problem with that argument is the failure to reasonably factor in time. If you take an expected value approach (quite reasonable in general) looking into the future you need to discount for time or everything goes to infinity.
As ben_w also hit on the time aspect, I ended up responded to him and you simultaneously here. [1]
[1] - https://news.ycombinator.com/item?id=48565321
While the point is valid, the upside still isn't practically infinite; humans intuitively do something close to hyperbolic discounting for future returns (I assume evolution did this because of the possibility of black swans but it isn't talking).
The easy version of this is to value a company over the next 20 years of profits. The first half of that time horizon, I don't see much changing at all outside of Starlink subscribers. They've only got ~5 Mars launch windows in any given decade, and while their design philosophy is great for LEO where launch windows are daily, launching rockets to see what fails is exactly wrong for Mars.
OTOH, now to 2036, there's at least a chance AI goes boom. Perhaps even one of Musk's AI. Great for the overall economy (at least right up until it changes the economy so much that capitalism looks as obsolete as feudalism looks today, which may or may not be in that horizon), but even if it's the next industrial revolution, there's so much competition there that I expect everyone's profit margins to be almost nothing. And conversely, if any one of them does appear to be massively dominant, in comes an anti-trust case (who remembers Standard Oil?)
Same with rockets, though a "rocket boom" is perhaps an unfortunate phrasing. He finds a way to make more money than expected, just like he's already doing with Starlink? Great. Industrial espionage is a thing, and even if it wasn't, China's not got a problem with the state funding businesses to simply re-discover everyone else's secret sauce. US companies and residents may find themselves restricted from using the fruit of that work, but most of the world will just buy what's cheapest.
Still, as I said, your point is valid.
Starship has already made it to orbit multiple times. It's likely that it will be ready to go operational in well under a decade. As we're speaking of upside potential, let's just assume the aspirational goals are hit - and you end up seeing $10-$20/kg to space. You've just revolutionized not only space, but all transport on Earth as those prices are less than most terrestrial shipping options, and it can be delivered anywhere on Earth in a matter of minutes.
The implications of that are quite difficult to even fathom because again this is one of those things where I feel like we're trying to predict the economics of the internet before the first cable had been laid. But the one thing that can be very safely said is that the impact would be revolutionary.
And that is completely ignoring space when the whole point of Starship is being able to get 100+ tons to the Moon or Mars, which again revolutionizes those domains first in terms of size, but vastly more importantly - in terms of cost as well. This opens the door to viable private development on the Moon and Mars (to say nothing of space itself). And again the implications of this are very difficult to predict, but it would also certainly be revolutionary.
---
The competition issue you've brought is, in my opinion, one of the most viable concerns. If America can get stuff to orbit for $20/kg then it feels like China always finds a way to do it for $0.20/kg. But the reality is that SpaceX's secret sauce isn't all that secret. Their exact alloys, general construction specs, and more are all public knowledge. Many have tried, including with vast resources and wits behind them, and all have failed.
China is almost certainly going to be the #2 player and their Long March 9 [1] is their attempt to clone the Starship, but for now it's vaporware. And to actually get these things out onto the launch pad isn't the endgame, but just the beginning of the race. They are almost certainly years behind SpaceX, and the distance is likely increasing over time as SpaceX shows no signs of slowing down, and now has no funding constraints whatsoever.
[1] - https://en.wikipedia.org/wiki/Long_March_9
> Starship has already made it to orbit multiple times. It's likely that it will be ready to go operational in well under a decade. As we're speaking of upside potential, let's just assume the aspirational goals are hit - and you end up seeing $10-$20/kg to space. You've just revolutionized not only space, but all transport on Earth as those prices are less than most terrestrial shipping options, and it can be delivered anywhere on Earth in a matter of minutes.
Let's assume it reaches $10/kg (I don't think it will, but let's assume that).
A 20ft shipping container can carry about 24 tons; the cost of shipping it to the other side of the world is much less than $240,000; the numbers I see vary depending on details, but even high numbers are less than $6k, which is $0.25/kg, and even then half of it is port-side not the ship's own cost. Air freight may get more competition, but not all shipping. And it won't replace a lot of trucks because it's too big (how many places want 10 container loads worth of goods at the same time?)
Yes, point-to-point 45 minute deliveries worldwide are going to have their uses, but price alone isn't revolutionary in that particular way.
Passenger flight would be something I can't rule out, but multi-gee launches will limit the numbers, the inherent noise issues will limit the launch locations.
I do agree with you that the implications will be difficult to guess correctly; however this is true in both directions, not just the positives, e.g. network hardware is vital to the internet, Cisco makes that, yet here's the history of Cisco stocks, with 25 years from the peak of the dot.com bubble to when it next reached the same price: https://www.google.com/finance/beta/quote/CSCO:NASDAQ?window...
> And that is completely ignoring space when the whole point of Starship is being able to get 100+ tons to the Moon or Mars, which again revolutionizes those domains first in terms of size, but vastly more importantly - in terms of cost as well. This opens the door to viable private development on the Moon and Mars (to say nothing of space itself). And again the implications of this are very difficult to predict, but it would also certainly be revolutionary.
Eventually, yes. But not soon enough for me to want to price it in today. Remember, hyperbolic discounting, black swans, competition, other things I can do with the money while I wait, all that means when I price it I want to know what it can do in the next 20 years at most. Even if I'm being optimistic about timelines, it would still take 20 years for the Moon to get enough industry for SpaceX to be worth the current market cap. (Well, unless someone makes a von Neumann replicator, but if that happens then whoever makes the replicator wins, regardless of if it was launched by SpaceX or not). Mars is harder.
> China is almost certainly going to be the #2 player and their Long March 9 [1] is their attempt to clone the Starship, but for now it's vaporware
Until they have enough confidence to put it into a circular orbit rather than one which has it hitting the Indian Ocean every time just in case something fails, the same can be said of actual Starship.
> They are almost certainly years behind SpaceX, and the distance is likely increasing over time as SpaceX shows no signs of slowing down, and now has no funding constraints whatsoever.
I assume China's behind, but with their secrecy (let alone the language barrier) it's hard to tell.
But SpaceX definitely has funding constraints. GAAP is says about $60bn obligations, but all-in they want to spend several hundred billion over the next few years and do not have enough revenue to cover it, so even with the IPO they are at actual risk of running out of money.
This is why the IPO happened at all: they ran out of private investors willing to give them money.
Most of your arguments have been perfectly reasonable, but trying to argue about the value of a shipping tanker vs point to point minutes time transport is not. We're speaking of viable same day delivery anywhere in the world, or same hour in some cases. Obvious complete replacements would be air freight, military logistics, and much more. For stuff you're fine taking months, sea shipping will likely remain favorable.
Cisco is more comparable to something like NVidia in modern times in that their valuations was driven by success during a time that was already screaming bubble, and an expectation for that bubble to somehow never pop. By contrast the space industry hasn't even really begun yet, so the wager on SpaceX is more speculative in one way, but also much more likely to be stable in another.
China isn't likely behind - they are definitely far behind. They can't even settle on a design for their Starship. And similarly, I think one thing SpaceX has proven is that fail fast is the way forward. You can't just sit in a sim and send these things up. There's too many unknown unknowns that you can only learn by watching things go boom, a lot. There is no world where China just unveils their secret Starship clone and starts flying successful missions with it. Again that's something everybody tries (in no small part because it sucks when things worth tens of millions of dollars go boom), and always fails with.
SpaceX has never had any trouble raising funds when desired. I think the main reason they IPO'd is as a black swan hedge. We're at a major inflection point on many levels, and it's very conceivable that in 10 years the US' role in the global economy will have been greatly reduced. Obviously that $3 trillion valuation is largely funny money, but for now that funny money still functions well enough. But that's likely liminal. It's entirely possible, if not likely, that we're currently Japan circa 1989.
It's not about the "Access to resources. A sense of adventure. Learning."
Since you are right. All of these things are available on Earth for far cheaper, with better ROI and with far less risk of dying or getting early cancer from cosmic rays.
The real benefit is a commanding lead over geopolitical rivals in the domain of satellites and satellite platforms for hosting devices in orbit, thus securing an insurmountable national security advantage.
The only argument for space is because you're so set on capitalism that you'd rather destroy the earth than stop trying to continually grow profit. If you're that sick and deluded, then space gives you a way to escape the consequences of your own foolish hubris.
So, your argument for SpaceX is that they'll take physical systems that they've already tried to squeeze down, and squeeze them down nearly two orders of magnitude? What fundamental scientific discovery do you think is going to enable this? Or do you think that AI is magically going to do it for them?
That has nothing to do with how they have already reduced costs, and how they continue to do so. Before SpaceX, you launched a rocket to space once, and then you discarded it. Imagine how much a plane ticket would cost if each time a plane was flown once, it was then discarded.
The big revolution with SpaceX was meaningful reuse (I can get into the comparison with Space Shuttle if you're unfamiliar there). Landing and reusing rockets is something that Boeing et al thought was impossible from an economic point of view, and actually taunted SpaceX in their early years over it along the lines of 'Oh you're going for reuse. Yeah we researched and trialed that out a decade ago. The economics don't work. Cute to see you trying though.'
Their success there is what helped bring the costs way down. But there's still plenty that's not reused - in particular they currently only reuse the first stage (the big rocket that gets things off the ground initially) while discarding the second stage - the space-optimized payload delivery rocket. With Starship they're going for 100% rapid reuse. So you're looking at this absolutely massive 2 stage system where both parts will be able to be repeatedly and rapidly reused.
You really think cost is the main reason humanity hasn't expanded into space? I'm all for space exploration and learning from space but actual humans are quite squishy, like gravity, dislike radiation, and would need to take a lot of water with them just to visit a rock very indifferent to their existence.
Definitely. Humans expand. It's probably an evolutionary imperative in many of us. Africa was essentially an oasis and yet some of us decided to go make our way outward to inhospitable freezing ass places where the environment itself is just constantly trying to kill you in a million different ways. And we knew nothing about how to deal with it, at first.
Or take the early sea voyages. Not only were there endless worries about things that were ultimately nonissues like sea monsters or falling off the edges of the Earth, but there were endless very real dangers awaiting which we had no clue about or how to deal with - scurvy, rogue waves, and much more. In the aforementioned Age of Sail, it was just expected that a significant chunk of your crew would die. Yet somehow we pushed onward and outward.
And the infinite possibilities of space are going to absolutely dwarf all previous frontiers in terms of interest and potential.
You're much more optimistic about the concept of infinity than I am!
You mean in terms of us creating a system of gods and serfs?
When you invest in TSLA and SPCX, you don't invest in a car and a rocket company, you invest in Elon Musk. So revenue, assets, are irrelevant.
You basically buy an asset whose risk is tied to a man's life.
And not the healthiest man alive
General Electric was just fine when Thomas Edison died in 1931. And Apple was fine after the tragic early death of Steve Jobs.
Companies like this become bigger than their founders.
And I'm sure many on the Left would argue that Tesla and SpaceX would be healthier companies without Elon Musk.
Although I tend to differ on that, having owned $TSLA for a long time. Never bet against Elon!
The limits are when companies and institutions start to default on their loans. Or, in the case of governments, trigger hyperinflation by printing money to pay off the debt.
Of course it can collapse before that, but if it gets to that point it is guaranteed to collapse.
But that won't happen while the share price keeps going up no matter what. Borrow more, secured against the massive unissued equity, or issue more shares.
> BTC essential is worth $70k solely through the power of memes
Like everything else in finance...
Saying this is not to defend all sorts of crypto-bros. The economy, especially one overly focused on publicly traded companies like the Western, and especially the US economy, is a meme economy.
Coffee, flats, healthcare, military spending, etc., of comparable quality in the abstract East, cost multiples less than in the EU/USA because they and their currencies are weak on memes.
Cursor was my first hands-on experience with AI. I didn't know much about getting set up with specific providers via API, and Cursor made it easy to pick any model, ask a question about some code, and get a clear suggested answer easily viewable in the IDE with an 'accept' or 'reject' button. I think they answered this question well: "How do normal developers want to interact with AI?"
I moved away from Cursor when I noticed the responses from specific models were not as clean or accurate as when I'd prompt the models directly, which was something I didn't know how to do early on. I hypothesized that they had some boilerplate prompt sitting atop of my own, causing less precise or desirable results.
I would assume Cursor is still one of the best options for normal developers to get started with AI, but with Copilot forcing their foot in the door at many companies, I wasn't sure how well it would fare on its own. Being acquired by SpaceX should help, and I'll be interested to follow along and see how things develop.
> Being acquired by SpaceX should help
Why? (Genuinely curious, I would have assumed the opposite)
My opinion is that the publicity can only help Cursor. I don't necessarily think SpaceX would make Cursor better. Copilot (which I view as a direct competitor to Cursor) has a huge structural advantage. I have several friends in various American companies where Microsoft products are all they are allowed to use. They get "free" Copilot access as a part of their Microsoft plans. Developers aren't having Cursor placed right in front of them in the same way Copilot is, and from my experience, when developers have the choice to pick one, they pick Cursor. So, I just feel the SpaceX/xAI publicity could help Cursor get more visibility in these general American software companies more so than they could on their own.
These are the SEC filings that confirms the merger:
Announcement of Cursor acquisition to SpaceX
https://www.sec.gov/Archives/edgar/data/1181412/000162828026...
Details of Acquisition
https://www.sec.gov/Archives/edgar/data/1181412/000162828026...
Congrats to the Cursor team!
When I first saw the company built on top of vscode in such a crowded field way back at the end of 2022, I thought "forget having a moat, they are renting their castle from the invaders!" - I couldn't see how see how a single team could execute well enough to effectively muscle their way in between Microsoft and OpenAI, who at that point looked destined to control the developer ecosystem between GitHub, VsCode and the then-best coding models. I think it's easy to forget how insane this seemed even just a few years ago.
But every year since then they managed to simply ship a better product on the axis that mattered to the most users. And now they are sitting between a huge user base and a massive stream of valuable tokens, they can sell to SpaceX. Incredibly impressive.
I am out of the loop with Cursor. Did they make their own LLM model or do you mean they just capture what is being sent to other models and that is valuable?
They have a few in models that powered the app, including one that applied edits. Recently they also started fine tuning Kimi models under the “composer” brand, and Composer 2.5 is a very cost effective coding model. But I suspect that the real value is in the distribution they have, which is what I primarily meant.
they probably just stole all private codebases cursor ever saw. That + the ARR more than justifies the valuation even without any internal model.
In related news, I'm open to suggestions for coding agent harnesses.
Happy pi.dev user here, give it a try! I would say that's kind of the "vim experience" but for harnesses: has the minimum, if you want something more you extend it :)
Saying it gives "the minimum" is generous, it's pretty much useless out of the box. And did I say slow as well? I think pi is great if you're into spending your time managing your harness rather than using it. In that regard, it's more like neovim.
> .. if you're into spending your time managing your harness ..
as is always the answer when this comes up about anything agent related :
save yourself the time and have the agent do it. Don't spend hours on something you don't want to do that it can - that's the point here.
I'm wondering who's going to buy pi!
(Edit sorry forgetting names, I mean who's going to buy Earendil). Good luck to Armin, he's done some good stuff.
Earendil did! https://lucumr.pocoo.org/2026/5/24/pi-oss/
That sounds like a Thiel funded project, given the Tolkien name.. If so, I won't ever be using Pi.
edit: From first glance, it doesn't look like it. But I basically don't trust any tech company that takes to Tolkien naming conventions.
The Earendil guys are European and very sane. Zero sf vibes from them.
https://mariozechner.at/posts/2026-04-08-ive-sold-out/
Sounds like it is not related
Theres several open source options you can use.
The highest github stars one is called `zed` another one i've heard about is `Cline`
theirs also a few that yc backed ones:
`Void`
`Continue`
`PearAI`
For what its worth the non yc ones have way more github stars but im sure the yc ones are good too. I think `Continue` is the biggest yc one.
I've used Cline for a year now and very happy with it. Fits my workflow really well
continue is eol after being acquired by cursor/spacex https://www.continue.dev/
and PearAI was just a fork of Continue with not much changed other than the license and removing the words continue.
Also OpenCode and Kilo seem popular as well.
At Poolside we have pool (https://github.com/poolsideai/pool), it runs in a terminal and you can use it with any Agent Client Protocol compatible agent (https://agentclientprotocol.com/get-started/agents).
Since you've asked, I am working on one but it is super early days so I am just posting it here for feedback.
https://zot.im
The idea is to make it fully autonomous so it is not really something that is meant to be constantly prompted and it is unlikely to fit most workflows but the idea is to make something that fits the future - not the present.
I've been pretty damn happy with codex and vscode.
Between the codex app, cli, and vscode extension there are options for most ways of working
Codex UI is great. It just make sense for an AI tool.
Been very impressed by the codex app tbh
I am seriously thinking about adapting Kate with a fully opensource ai harness. it should be good enough for mac and linux for most devs. it already supports lsp server & has a established plugin infra so it should not really be any blockers. anybody wanna collab?
this class of spyware pretending to be ide makes me sick.
Nvim + Claude code. Or zed. Honestly basically every ide is adding agent harness features as their primary focus right now. Just throw a rock and you’ll hit something that will work for you. You can even just use emacs and have ai build harness features for you.
I'm very curious what coding agent interfaces other HN users are using.
I run a bunch of Claude / Codex TUIs + vim in terminal tabs on i3 workspaces on Linux whch I know isn't the most common.
https://github.com/can1357/oh-my-pi
If you're open to using remote Linux VM's and a web-based interface, I quite like exe.dev and Shelley.
Paseo with opencode backing.
pi is for real engineers
vim
Hm, surprised at all the Cursor hate here. Tab complete, at the quality they delivered, was a game changer back in the day.
And their current work on Composer is great. Composer is super fast and quality is decent. More competition in the model space always welcome.
"Back in the day" is the key idea here. It's the valuation that doesn't make sense in today's market. What does Cursor offer to be valued at $60 BILLION? Its models are alright but nowhere near SOTA, and pretty much superseded by open source local models at this point.
Composer 2.5 worked just as well for me as opus 4.6, and was faster and actually worked better for quick bugfixes and reporting back to support. And is much cheaper.
I'm not saying HN should be super supportive of everything, but the level of hate and complete loss of reality for a lot of people is quite sad to see, for a community of supposedly intelligent people.
I don't know if it's just me, but the thoughtfulness of of the average HN comment has seemed to decrease in the last few years. Feels more populist/Reddity the last couple years.
You haven't been paying close attention. This has been going on for about 15 years. I'd say we had 5 good years here before the quality began dropping precipitously. If you got here after about 2012, you're probably part of the problem.
I'm pretty sure it has. If you look at threads from a decade ago, they're completely different - much less dramatic, much higher quality, way less populist and full of manipulative language.
Use those downvote and flag buttons!
"Back in the day" was... a year ago
Can't remember the last time I used AI tab complete; back in the day indeed.
HN is a hellhole of either AI hate or hype. The fringes on either side are the loudest.
Nothing wrong with Cursor but $60B, wow. How many of these deals in 2025, 2026 will be worth nothing in 5 years? Seems like everything is just desperation and less like long term strategy.
Wow everyone on HN seems very out of touch with Cursor.
Their UX on their agents only window is quite good and on par if not better than Codex and Claude desktop apps and it still has quite a good bit of subsidization especially with Composer 2.5 being on par with Sonnet at the very least. They're also growing tremendously this year to 4B ARR recently on track to go higher before EOY.
Exactly. People here still think that Cursor is just a vscode fork while it has positioned itself as agent orchestration console that happens to have an editor you can switch to.
Claude Code has captured so much mind share, it doesn’t even matter. Cursor is absolutely cooked.
I don’t really understand what the value prop of cursor is, it must be the data and models. These days programmable editors like neovim and emacs have a huge advantage. I’ve had ai create several custom plugins to have my editor do whatever I can think of. Just ask Claude code, hey I want to do x, y, z, it spits out some lua and I have a new capability. I don’t know why anyone would want to be limited by an extension interface at this point.
>I don’t really understand what the value prop of cursor is, it must be the data and models.
The simpler answer is that there is almost no value outside of buying some customers.
As you've proven to yourself the engineering work is doable on your own.
I've made my own agent and wired it to emacs via ACP... 60 billion in value, ok... sure...
https://agentclientprotocol.com/get-started/clients
For what it's worth, this was effectively announced months ago, and at this valuation.
But they (SpaceX) could have backed out of the deal at any given time as they had the option to (and be required to pay the 10B break up fee). Nobody knew what would happen at the time.
This announcement is a definitive agreement of the acquisition at that $60B valuation.
Sure, not saying there's no news hook here.
Yes, that's what was said basically. The pedantry is not adding anything to the conversation.
I don't think you can even say that without reading the actual announcement. [0]
Was the acquisition "effectively announced months ago"? or was it the right to acquire the company or pay $10 billion that was announced months ago?
It would be equally relevant if it went the other way. But clearly both of you are confused on what was said they would do in the future (which that was the announcement months ago) vs what are doing today.
[0] https://www.reuters.com/technology/spacex-says-it-has-option...
Besides now paying 60 billion for a fork of VSCode, it seems SpaceX meme stock style money raised from the IPO, is all gone in one week :-)
IPO proceeds after greenshoe: $85.7B
Major disclosed cash / debt-related commitments:
- Take out Bridge loan tied to X/xAI debt repayment: $20.0B
- Take out EchoStar debt payoff / cash component: up to $8.5B
- Take out EchoStar debt-service funding: up to $3.0B
- Take out AI infrastructure lease commitments: $20.2B
Subtotal of major disclosed commitments: $51.7B
Rough remaining cash before other costs :-)): $34.0B
Lets now talk about buying Tesla, doing Quantum and building a Dyson Sphere and do another round?
They are paying for the installed based. The users mainly. And also the team.
Looking at the comments on this thread, those users/installed base will be gone by the afternoon...
Does anyone here think Cursor is overvalued? It's just packaging up what already exists, it has no moat or IP.
They're getting paid in extremely overvalued stock, so maybe it balances out.
This is not really a diss on SpaceX either because a lot of IPOs go through an immediate pop and then 1-2 years of doldrums as lockouts expire and promises aren't quite delivered.
Nobody knows what 60 billion in SpaceX stock today will be worth when Cursor insiders finally get to sell (at least a year from now, after other SpaceX insiders have started selling).
Source on the lockup for Cursor insiders?
Generally this is how liquidity works. Their employees will have a six or twelve month lock up (six being most common).
Investors in certrtain rounds (or sizes) tend to have no lockup, whereas later stages have a six month. Alternatively, I've reviewed agreements where the lockup is based on minimum market cap, but I've only seen that a couple of times.
Ok everyone saying this is how it works but where’s the proof? SpaceX has a 7 day lockup for some people which is abnormal. So clearly the way it’s done isn’t in fact how it’s always done.
This is just a general practice that always happens when paying in stock. It's to prevent a massive dump the next day which would tank the share price 'artificially'. Again, rich people's rules.
Ok everyone saying this is how it works but where’s the proof? SpaceX has a 7 day lockup for some people which is abnormal. So clearly the way it’s done isn’t in fact how it’s always done.
https://www.sec.gov/reports/rule-144-selling-restricted-cont...
> Holding Period. Before you may sell any restricted securities in the marketplace, you must hold them for a certain period of time. If the company that issued the securities is a “reporting company” in that it is subject to the reporting requirements of the Securities Exchange Act of 1934, then you must hold the securities for at least six months
Are the securities restricted though? I don’t think these would be?
Almost guaranteed to be. The only other way would be to have issued them during the transaction with the Sec. There was no mention of that. 60billion of issued shares would have been mentioned. It's non trivial amount.
What’s the lockup period in this case?
Um, I honestly don't know for sure, if I'm not mistaken SpaceX has some weird staggered release schedule for employees and early investors, I guess based on their stock dates.
> SpaceX has a 7 day lockup for some people which is abnormal
You might be referring to staff members who have shares ? Their shares are not restricted securities as far as I know, but their internal company policy might affect those, but I'm not 100% certain on that.
There are many tranches here. Some friends and family got to buy day one IPO with no restrictions. Then some employees get a rolling release starting June 30.
That's interesting actually, and good to hear. I fully believe those employees deserve their full share.
It can be inferred because it’s how things tend to work.
Ok everyone saying this is how it works but where’s the proof? SpaceX has a 7 day lockup for some people which is abnormal. So clearly the way it’s done isn’t in fact how it’s always done.
Well, the people who bought the SpaceX IPO essentially footed the entire bill here. And they might still make money on it, depending on how the stock goes from here on. I don't see anyone who could lose here, even if the bubble bursts, apart from the Cursor people. And they are likely still going to make a huge amount of money.
> Well, the people who bought the SpaceX IPO essentially footed the entire bill here.
It's hard to say that they footed the bill here, but they basically gave SpaceX a number to say "well our stock went IPO and it's at this price, so here's 60B at this price"
A good tactic from SpaceX as after the inital surge of a big IPO, the stock price usually comes down and finds it's correct balance, which is usually always lower. So if they had of waited the 'cooling off' period of a year for example, and the stock price went down to it's 'correct' valuation, then they would have had to issue a higher number of stocks.. At least that's my thinking, but I'm terrible with money.
No, look a Composoer 2, it stands out starkly on its own in the pareto frontier on low cast and fast models.
Composer 2.5 was a huge leap with minimal compute from xAI.
They can compete with OpenAI and anthropic with xAI scale compute. They have a top notch model team and incredible training data and huge enterprise costumer contracts.
my employer (one of those huge contracts) dropped cursor in favor of claude and i don’t think this is true at all
while we had it i used cursor for probably eight months as my main ide (i did really like the interface for embedding code in prompts!) but had no problems switching to claude code. i asked around, and i truly don’t know a single coworker who misses cursor even a little bit.
My experience mirrors this as well.
I was fully in on Cursor for a good chunk of last year, using Composer + Gemini Pro (via Copilot / GH integration). I really enjoyed Cursor's tab completion capabilities, but when Sonnet and Opus started getting particularly good for me (think for me it was around 4.5), I swapped over to Zed + claude code in the integrated terminal. I've found that after a bit, I haven't ended up missing the tab completion. I've been perfectly fine with just LSP + claude always open. I don't miss Cursor. All my colleagues are on claude code with half of us also using Zed.
Composer 2.5 Fast is particularly good.
For someone who is new to agentic code or is generally somewhat junior, Cursor is very easy to get started with and is generally fairly frustration-free.
I use a cheap $20 subscription mostly for occasional use of Opus and Composer.
SpaceX made a smart move here. Someone else should have really seen the opportunity and bought them.
Composer 2.5 is just a repackaged chinese model, Kimi IIRC
How the hell is an IDE a "pareto frontier"? Even if, say composer 2.5 is a huge leap forward, that doesn't mean IntelliJ or Vim or Emacs or Codex got worse.
IDE improvements are not a zero sum game.
In a competitive environment, it's precisely that. An improvement in product A takes customers from B and C.
Nope. Vim and Neovim users can use the cusor-agent cli for agentic stuff and there prefered editing tool for editing. All the major providers have a cli specific version these days. Probably because folks actually didn't want to actually use the cursor Gui and once Claude code came along those folks jumped ship and went full cli again
This is delusional. Composer 2.5 is trash compared even to haiku let alone opus.
This doesn't match the empirical observations of a lot of people I'd trust more than you, and putting it below Haiku immediately makes it extra sus (Sonnet would have been the credibility preserving comparison).
Hmmm. Not in my experience. I don’t think it can be compared to Haiku, maybe sonnet levels? It’s obviously not Opus and never was intended for that use case. I use it quite a bit and it works well and is extremely fast for the tasks it was built for
It doesn’t tell anything about the valuation, but I prefer it over Claude Code, and I even stopped using JetBrains IDEs because of it.
Vs Claude Code: I like the option to change the models, as I often prefer ChatGPT or Composer to Opus. I have a slight preference towards TUI, but not so strong to drop the models.
Vs JetBrains. I really love JetBrains but the tab complete just works so well for me.
why u prefer cursor over claude code?
its not just the models, their auto complete is actually really good. when you make a change it will give you "tab to next" which makes refactors super easy.
composer 2.5 is also a very decent model, it go 90% of my AI tasks using it now.
So 60B dollars and your primary reason it’s worth that is “tab to next” autocomplete?
Scaled to every developer in the world? Yeah, that's productivity.
I'm not sure it's worth the price tag for developers though, I mean resharper promised similar, delivered half, bought by millions and still don't think Jetbrains, even with its other good tool suit is valued 60B.
Investing in companies is not about what they are delivering right now, it is about what you think they can deliver in the future.
It is not like purchasing soap in the supermarket.
I just laughed out loud reading this
Hyperfocused edit assist is awesome. I’m the magician the assist is the magic wand.
Maybe the real Mythos secret is that it found out it could fit on a GPT2 size model
Not to mention VSCode has that too
Yes, that's why Cursor was very popular when actually reading model output paragraph by paragraph was still the way you used them. That's no longer the case, their use has cratered, and in fact they have been disintermediated by their model vendors, leaving an empty shell.
I guess that fits into the Musk empire because "empty shell" increasingly describes his companies.
Tesla is a car company that doesn't want to make cars. And xAI is an AI foundation model company that actually is a data center REIT...
I tried it a few times and was not convinced by the autocomplete.
I found it less effective than free copilot autocomplete on vanilla VSCode.
composer 2.5 is literally just a fine-tuned kimi model, and the autocomplete is exteremly meh.
The only kind of AI I want in my editor is an autocomplete, but this isn't very magical to non-programmers (their TAM) or all that valuable (you can't charge thousands), they bought Supermaven and basically killed it, I'm not sure how you think tab is really good, I've not been impressed when I played around with it.
Yeah, is that a bad thing? They did some really interesting stuff when fine tuning.
https://cursor.com/blog/composer-2-5#targeted-rl-with-textua...
What tab is better?
I'm honestly not up to date on all the FIM autocomplete at the moment. However, autocomplete getting good is something I look forward too. I believe its mostly a problem of latency. You needs <200ms (and accurate) for it to not feel annoying, which is hard to do and use a larger more capable model.
Unfortunately it doesn't seem like this is a problem companies are looking to tackle because it doesn't really make you any money or impress investors. Inceptionlabs.ai seems to be making attempts, but I haven't tried. They are apparently in the Zed editor, something else I haven't used.
https://www.inceptionlabs.ai/blog/introducing-mercury-edit-2
Cursor was nice when I was still meticulously hand coding my stack, fantastic autocomplete. With today's top models, I barely write code myself, just review commits. Cursor eats Opus credits like there is no tomorrow. Composer has been a net negative in my experience. All in on Codex with GPT 5.5 on high using /fast.
IMO composer 2.5 is pretty good, and the $20 Cursor plan gives you _way_ more tokens than Codex/Claude code/Antigravity
I seem to get a lot more tokens for my $ with the $200 plan from Anthropic or OpenAI than I do from the $200 plan from Cursor.
Lately I use Cursor with DeepSeek API, and OpenAI subscription through their Codex App.
Not my experience. Maybe an hour of top model use on the $20 plan, then Composer 2.5 which needs constant hand holding.
Only cursor employees could continue to spread this idea that composer 2.5 isn’t total garbage.
Cursor, from my companies perspective at least, seems to be handling charming leadership to get enterprise AI contracts in place, compared to the alternatives. That's feeling like the moat from my first-hand experience. Easy single contract that covers a lot of AI cases that management wants to say they have in place.
Yes. I think Cursor is overvalued, but not to the extent of SpaceX
Maybe it's just the US dollar that is overvalued
I get the joke, but if anything it's that there's too much money sloshing around.
That would explain the top tier of meme stocks, but counterpoint: what about our salaries?
Workers are criminally underpaid
Could not say on the valuation front. The one thing that is obvious is the trash talkers love to point out how horrible it is but I have yet to find a compelling replacement. Obviously Claude Code works but at least for me I never got along with the CLI workflow very well and sure I could use vs code with the extension for Claude but then I lose tab autocompletes which I actually like. I have yet to see anyone build a derivative model like composer 2 that is quick, cheap and has higher reliability on tool call use. Again I don’t know on valuation but it’s pretty impressive how far they have come. I look at Jetbrains and at least from an AI perspective they have been left in the dust.
Using Cursor with Jetbrains via ACP isn't that different than using within the Cursor IDE. You still get Agent/Plan/Ask modes with model selectors.
I'd recommend installing the Cursor IDE, then using it to install Cursor CLI (it's easier to keep things up to date this way), then setting up your Jetbrains IDE to point to the Cursor CLI via ACP integration.
Cursor's moat is that it is a virus that infects organizations through shared skills, hooks, agents, etc.. Once one person uses it and infects the repo everyone else starts using it.
Know a few companies that have moved from it fully to Claude. It’s still early so the moving cost is low and Claude Cowork is something non-tech employees can make use of much easier than Cursor. I really don’t see what Curor’s value is longer term. Why pay a middle man?
Maybe, but migration from/to anything is so much easier now with Agents.
I don't. Cursor being a man in the middle between coders and other people models for so long, has so much more training data than anyone else in the world.
There is obvious proxy to the amount of training data - revenue. And I think anthropic is way ahead of them.
There's a plausible synergy in it for xAI though. Access to reams of training data for a company whose marginal cost of compute is very very low, and that they can use as a channel to push Grok. I don't think it's worth anywhere near this much to anyone else, but to xAI it's at least possible.
None of that really matters.
What matters is that this has enough "future story value" to keep the few investors invested... allowing for the planned index funds to buy into the overvalued stock & allowing for the largest heist in the history of money.
It's become pure hype and drama on the global stock market stage.
Transactions that happen out in the open, with the consent of all parties are not a heist.
Money is neither created nor destroyed in the markets. It's exchanged.
If Musk is worth a trillion while not generating a trillion of value, then it's a heist.
heist: "an elaborate, meticulously planned theft, typically targeting highly valuable items, large sums of money, or financial institutions like banks or museums"
Scam maybe.
Moat is whatever thing is stopping the next guy from simply drinking your milkshake. People conflate "I could smash this out in a weekend" with "and therefore could also build a multi-billion dollar revenue stream in [counted in months]".
I do think this has had its day. From what I remember, Cursor was useful back in the day when you coded in an IDE and wanted to read code while you baby stepped through incremental changes with an AI. I'm tempted to put /sarc around this but not really...
They have a surprising amount of enterprise revenue and mindshare, of which xAI has literally none.
Sounds like what people here said 20 years ago about Google buying YouTube, or 10 years ago about Facebook buying Instagram - companies with no moats and huge infrastructure costs.
To paraphrase, the biggest trick the devil pulled is convincing founders they need a moat.
In terms of whether they’re overvalued: probably. But any valuation should also take into account the value they have to x.ai (also under SpaceX) as a source of training data for coding models.
I like Cursor but that valuation is a hard sell.
Their valuation should be very closely tied to how how many tokens it takes to get from Void to Cursor.
If those values diverge by much, something is rotten in the state of Denmark.
Cursor Remote Agents are important to our AI orchestration layer. It's possible that Claude can do this directly but Cursor Remote Agents made this laughably easy.
I guess they're getting bought because they had access to a lot of codebases from a lot of companies, and perhaps there's something to mine in those logs...
Composer is very good, and these days after heavily using CC, Pi and Open code I am back with Cursor. "No moat or IP" is underrating it a lot.
CC on Deepseek is a moat.
they propably have a lot of training data from their users, which might be useful for SpaceX which has a lot of compute
It's not true it has zero moat or IP (they have their own LLM and it is useful), but it is indeed way over-valued.
Grok needs a coding environment play to match Claude Code - that's what this is.
And AI companies are not short of capital.
Thats grok build, which is pretty nice and has compatibility with Claude Code. They just need better models, that will be released soon.
They have the conversation history of every person that has used their product. That's worth something.
Every single one?
So is 'Space'X. They fit perfectly
obivously is overvalued
That's a lot of money for a buggy product that is at best slightly better than its competitors.
Well, when your stock is massively overpriced it's a smart time to buy stuff with it.
I dont understand if AI is so good at code why are people buying software? couldn't musk just generate his own IDE with the spacex brand behind it?
AI is good at code but it can't one-shot whole platforms. I don’t think we'll be in the "give me a new Windows XP but Unix core" range anytime soon. You still need time and guidance/ideas.
not totally true anymore. check out https://programbench.com/
new models like Fable were scoring 30%. wouldnt be suprised if very soon we eclipse 50%
I think they are buying data and customers.
My honest suspicion is that Musk will focus more and more on AI (and less on space) because he sees it as a path towards his immortality. I expect AI models trained on him combined with millions or billions spent lobbying to allow an AI to own and direct a company. I know this sounds like poorly written sci-fi but I will be only disappointed - not surprised - if post death AI Elon is the richest entity on earth.
AI + Optimus + desire to explore the universe is such a cool concept. I think he'd love to send his consciousness out after all this bickering on earth.
Honestly, the more that we get humans away from squabbling over the same patch of desert the better off we'd probably be. Everyone seems to think they can spend money better than everyone else.
Absolutely unhinged take. Most people just want healthcare, affordable groceries, and a vacation dude. I think Musk has a better chance of being guillotined in the next 20 years than he does sending his consciousness into space.
That'll be true of anyone no matter how much money they have, there's a reason those lines are being sold for votes. There's no end to them and people will always want more, it's not a true goal to work for. Somehow enough people don't understand basic economics enough that they really think taking Elon's "money" (estimated net worth with a million asterisks) will change any of the above for everyone.
Maybe Peter Thiel and Elon Musk will meet again cryogenically preserved ;-) https://fortune.com/2023/05/04/peter-thiel-cryonics-cryogeni...
They sold at the top. My entire TEAM was weaned off cursor in the last year. New setup - 50% of them do (Codex Desktop, Claude Desktop) + Zed for IDE and the other 50% use (claude code + codex cli) on cmux -- a ghostty based terminal that adds some bells and whistles, literally for notifications when claude is done.
IMHO, the codex desktop app is very powerful for development + testing given it can easily control the computer.
paying 60bn for a dev team that wrapped vs is insane.
How about paying 60 billion for a bunch of enterprise contracts?
I suspect these won't be as sticky as the contracts you're thinking of. Oracle and Microsoft can run on that because their products aren't compatible with anything else so migrating is a huge pain.
Cursor doesn't really have that. We've got it at $DAYJOB but it's not even the only one, I can also use Zed or Codex or Claude or probably a half dozen other things I don't even know about.
I suspect a lot of companies have that right now because the market for AI is so volatile, and in the near future will trim that down to a couple of tools and Cursor doesn't have much to keep them at the top of the list imo.
I also wouldn't pay 60 billion for a bunch of enterprise contracts that have to compete against Microsoft. No one is dropping Microsoft as a vendor, and they have the ability to up the prices for stuff people need like Office and use that to make Copilot free. If times get tough and companies need to cut costs, it's a lot easier to part ways with Cursor than Microsoft.
Or paying ~300 million virtual slips of paper that the market says are currently worth $60 billion.
About 4x what they've spent on Starship development so far.
They're not a space company any more. They're just part of the AI bubble.
Not bad for a VS Code fork and a Chinese LLM fine tune.
Are people still using Cursor? I haven't in at least a year but perhaps I'm in the minority.
Yes. I like it for small fixes, specifying exact code ranges i want touched, and for asking questions because it links me to the code so I can read it.
I use Claude more for greenfield feature building where I dont need to surgically dig in and view existing code specifics.
Yup still got it due to paying for a year up front, its still very decent, especially the newer composer 2.5 model, it's cheap and has a ton of usage included so works well as a general day to day tool.
It’s the best exit for Cursor. There’s not much of a path for Cursor besides getting acquired. Cursor is a fork of VSCode. How much improvement can you really make? Cursor’s own model is based on a Chinese model. OpenAI and Claude are SOTA for coding. The only selling point for Cursor’s model is that it’s cheap.
This is the best outcome for Cursor.
Absolutely. Imagine forking VSCode and then selling it for $60 BILLION in a few years. And right as you lose your competitive advantage to foundation providers. Amazing.
I don't understand. What SpaceX gets if they stop developing the product - as your comment sounds like it is a dead end? There would be then zero sense to buy it?
cursor feels so 2025 to me guys. these days zed is just way better for my macbook battery and with acp can talk literally to my installed claude code and codex CLI tools, plus their own and custom providers ontop. I was kind of a decade of a vscode user and always just stayed through the evolutions until cursor, but at some point I just need a lean fast editor+lsp combo, git included and a chat pane next to it that uses my real subs underneath easily. (also: codex-cli can spawn and manage subagents and _resume_ them, acting like a real manager).
could be only me though, but longer interactions over days makes my codex gui app grind to a crawl and cursor was not only expensive with opus via api costs but also heating my room a lot. now I have a dozen zed instances open all crunching along with LLMs barely noticeable on system load (except the occasianal testuite runs but thats expected).
If you've ever used Cursor, login, make sure privacy setting is on and contact them to say you want all your data deleted if there is some stored. Then delete your account.
Anyone recommend an alternatives? For no particular reason I’m canceling my cursor subscription today.
If you want something price-competitive with composer 2.5, deepseek pro is very cost-effective. Just rig it up in opencode via openrouter.
Don't use OpenRouter, the company is a shitty middle man. Use DeepSeek API directly.
Composer 2.5 is very heavily discounted with a Cursor subscription. You effectively pay 2% of the API price of Composer 2.5 Fast with a subscription.
Does Deepseek offer any discounted tokens subscription like that?
Everyone I work with who used Cursor stopped using Cursor when Claude Code came along. They're back to their regular IDE when the need to read code, or they just review it at PR time. I never used Cursor, but Zed is my favorite editor with an agent. It can use Claude Code, among other CLIs, via ACP, so you can use rolling subscription tokens, or it can use OpenRouter or others if you want a broad spectrum of models. And it's crazy fast. It used to be that Copilot Pro was the best deal on agentic coding with several models from several vendors available, but they've really nerfed it, with uselessly restrictive token budgets and only older models are now available from the major labs. These days, might as well just have a Claude or Codex subscription and use the CLI with ACP in whatever editor you prefer.
I've been really liking the Cline extension in VS Code
Opencode with an Opencode Go subscription is a great deal.
Second this
Completely valid way to sign out!
Check out zed
I wonder where they will take this, if they'll use the Cursor team to help make Grok Build (which is not just a tool like Claude Code, but an actual Grok model too) more refined for programming? Would make sense to me, and in turn also provide Cursor with more compute they can use.
> For the past few months, SpaceXAI has been jointly training a model with Cursor, which will be released in Cursor and Grok Build soon.
https://x.com/SpaceX/status/2066873915717136548
It also provides xAI with a pre-existing enterprise distribution channel. At the end of the day, distribution is equally as important as the underlying product itself and in some cases is even more critical.
I think that's another thing, while I would like to test Grok more, the Grok AI plans are very generic and not tailored specifically for programming, which is frustrating because I get maybe 8 hours out of Grok for $40 for an entire month, I do wonder if they offered a "Grok Build only plan" if it would actually give me access to more compute. Maybe they intend on making it through Cursor.
I do hope that Cursor doesn't remove any of its current model offerings, and just offers Grok Build in addition to what they already offer, in my opinion unless most of their clients "switch" to Grok (like metrics show they're mostly using Grok vs other models), it would make more sense.
No company that uses Cursor now is going to be ok with using xAI models.
Not necessarily.
Enterprise AI adoption has reached a point now where FinOps matter, and a harness platform story with a discounted underlying model can be enticing for a number of organizations.
I've seen Gemini land well in a F100 well known for their AI hardware story for that reason, and Alibaba's leadership canned the OSS minded Qwen team in order to build a similar commercial minded approach as well.
At least in cybersecurity, we're also reaching a point where the harness is starting to matter more than the underlying foundation models, and building a harness/bedrock style story while discounting a specific model can play well in upper market deals.
What makes you think so? Is it only because you dont like Musk or do you have some insight into all companies using Cursor you want to share with us? Even if you dont like Musk you should realise that others may not share your sentiments and/or may have similar feelings regarding SamA, DarioA or any of the other CEOs in this field.
Enterprise customers care about things like guardrails and data safety. xAI has always been anti guardrails, and who knows if you can trust them with your data.
At least in the F1000 RFPs I've seen and the decisionmakers I've chatted with, when they talk about AI guardrails what they mean is generic API (eg. can we rate limit, block connections, RBAC/ABAC capabilities, etc) and Data Security (eg. ZDR, encryption at rest/transit, controlled access) controls.
There is a recognition that foundation models and tools leveraging them will introduce some degree of nondeterminism, so the best way to solve that is to leverage preexisting best practice that is used to reduce lateral movement risk by humans (who are similarly nondeterministic in nature).
My company’s security team is very much “no proprietary data or information can be used to train a model”, I just don’t know how you can validate or trust that they aren’t doing just that.
ZDRs, spot audits, and the fact that salespeople can be held personally liable both financially and even criminally for fraud if they sold a contract with a ZDR that was not honored.
They'll likely just take all the various enterprise contracts to help inflate SpaceX earnings, that's the only reason why you'd buy this if you were Musk. Need to help keep the lie going as long as you're still alive.
Cursor's value add as a developer seems much slimmer than the 60bn price tag justifies, but I guess they have a lot of data from the non private usage which bumps the value up?
The product itself is practically a vscode wrapper with Agent implementation and K2.5 forked model (composer).
I’d rather buy 60 “overpriced” Instagrams in 2012, or 3 overpriced WhatsApps in 2014 or 1.5 overpriced Twitters in 2022. I can’t tell what’s a bubble and how money works anymore.
https://news.ycombinator.com/item?id=47855293
Initial announcement back in April
Not sure how this closes the gap to Anthropic and OpenAI for xAI. Is there a play that I am overlooking?
If this acquisition goes through the only winner here is Cursor, especially since CC and Codex are chipping away at Cursor very hard!
$60b is genuinely insane. Very high from a P/S ratio perspective, and for a product with arguably no defensible moat.
Congrats to the Cursor team though... One of the most crazy exit stories ever – 4 years to a $60b buyout. Damn.
I like Cursor as a product but the add on cost of $0.25/M tokens is just too expensive on top of the models.
Yeah, I don't understand how this business model continues to produce revenue.
Revenue? In that industry?
Charging a percentage (like OpenRouter) would be much more fair IMO as smart-planner-cheap-subagents becomes more common.
They didn't wait a full week before dumping shit in the investors.
Why would spaceX want to double down on AI after the pain xAI is giving them with no good models and no use for the hardware?
As I understand, the AI part makes $2B a month now (from Anthropic and Google).
The AI/compute revenue is higher than what the space department makes.
SpaceX is 90% xAI and 10% space. It's no longer a space exploration company, it's an AI company (with a sub-par product). Insane (and dumb) valuation, kind of like Tesla.
I wonder how much zed industries is being valued at.
For those who aren't aware, Cursor sports one of the best LLM harnesses for coding. The app itself is annoying to use compared to their CLI counterparts, but the harness is widely recognized as the best in the business, or very close. Buying that harness makes a lot of sense considering the cash Musk invested in Grok. He's clearly trying to play with the big boys and grab a chunk of the LLM-assisted dev market.
How does this get a Starship to land on Mars or coast-to-coast full self driving? $60,000,000,000 towards one of those goals would have checked-off one of those boxes.
It's an all stock purchase, so closer to a merger than spending $60B
When does the Tesla acquisition get announced?
Within the month. They'll probably try to do it sooner while the stock is at a high.
Or do it when the prices are starting to show weakness
I'm still waiting for the real news to drop- in the next 6 months we're going to start hearing some big moves from Space X AI. Early this year they lost pretty much all their leadership, it's very clear they failed to keep up with the frontier models and Musk has essentially given up for now - renting out their compute to Anthropic and Google. But that's not sustainable, everything they say about their IPO is that AI is the core value driver. So at some point Musk is going to have some decisions to make about who he brings in to drive that. I imagine once they get that person in and start building a team around them the deals with Anthropic and Google will be ended.
I guess the cursor guys will be happy because they got their pay day, but I'd be very aware if I were them that their future is at the whim of whoever Musk appoints and it's difficult to tell who that would be right now.
I guess now is the time to take bets, so I'm going to bet an early OpenAI employee like Sutskever gets the job and they acquihire him in. Here's a bit of a laugh - at this stock price Musk could probably tempt Demis to come over, that would be wild.
Why is it not sustainable? Right now it appears to be a better business than selling access to models.
It's a good business. Maybe even a great business. But it's not going to justify a valuation like Space X. In the same way that Tesla has slowly become less competitive in automotive, I don't think it's sensible to assume Space X will have some durable edge in building data centres, especially when basically everything going into those data centres are commodities. And if it is just a neo cloud, then you have to contend with the pro-cyclical nature of that. It's also just clearly not their plan, they're not promising to be a neo cloud. They're promising to own the full stack.
Props to Elon: he IPOs a space based telecom company wrapped in a bunch of garbage companies (xAI et al) as a meme stock and then uses those Elon bucks to go and buy a company to make his struggling ai ambitions relevant.
The formula is: create paper wealth, use that paper wealth to buy something else, and rinse and repeat.
The rich just live in a different part of the multiverse than the rest of us.
Wow they are using 80% of what they raised 4 days ago to buy an IDE. Absolutely incredible.
Per latest reporting, Cursor's current annual revenue rate is $4 billion.
So SpaceX is buying 15 years of future cashflows, assuming no change in revenue. Which is bad assumption given past growth was gigantic.
They are also buying 300-400 employees with proven record of training good coding models.
Their expertise will be used to improve Grok Build coding agent.
Also, they're using stock, not cash, so effectively they doubled the amount of money raised.
>Their expertise will be used to improve Grok Build coding agent.
Is Grok not a toxic enough brand (by association with Musk) that people who would use Cursor wouldn't avoid Grok?
Like, the assumption seems to be that all the goodwill that Cursor users have towards Cursor will now apply automatically to Grok, which seems like a pretty significant leap.
I've been teaching myself physics lately and have found Grok to be one of the best both at coming to a correct answer and helping me to understand how to get it myself. It also seems a lot better than other models at saying "I don't know" or pointing out when my question doesn't make sense.
I bet any flagship model would do as well if you prompted it with how it should do it.
Comparing grok vs Gemini vs GPT vs Sonnet is like comparing mid-high end CPUs. They're all about as good as one another for most work.
Grok has one of the best reasoning and halucination benchmark scores.
[flagged]
I'm just trying to get it to help me learn physics. If my topic of interest shifts to mid-20th century European history I'll keep what you said in mind.
You're supporting a company that makes a model that has been intentionally directed to think that white genocide is a major problem today.
You do you, but that's a very morally implicating choice you're making.
That's your misguided opinion masquerading as fact.
In no way is he a nazi or any of the other ad-hominem attacks y'all throw here.
You'll probably point to one instance of an awkward gesture, like Elon isn't awkward. Clearly hearing him talk, he's not a nazi or racist.
That wasn't really an ad hominem attack.
If you're going to use the model to learn history you're going to learn the version of history that the model teaches you. A little bit of digging around grokpedia should give you some idea of what that model thinks
Was in reference to GP's "neo-Nazi" comment and similarly expressed HN users here.
But you may be seeing your bias if you think grokipedia is wrong.
Probably being used to leftist editors on wikipedia would do that.
Or maybe it's somewhere in the middle for some events. You can always validate sources and determine for yourself.
I'm a heavy Cursor user, I spend hundreds of dollars per month in overages above my $200 subscription, and I'll be switching away. I have zero interest in Musk's AI.
Companies using cursor could not care less about the CEO's ideology if they tried. Individuals may, but they don't matter in this context.
The CEOs ideology matters due to the fact it impacts the product design. The reason people don't want to use Grok is because it's bad, and it's bad because the team behind grok have to spend cycles crowbarring in far right white genocide conspiracy theories so that it doesn't embarrass their boss on twitter. One of the things we learned with Anthropic is that you have a lot more success being focused on the core product - coding agent, than trying to do that and chase internet chatbot users.
As someone that trains LLMs, even if grok does have a "avoid ""woke""" fine-tuning, adding that needs a few thousand examples SFT and a system prompt. It costs nothing extra to do it to coding agents and is not the reason why grok sucks at coding. It's just not in the same league in general - it's 0.5T parameters only and not trained specifically for coding at all.
Even if the way they are doing it did damage coding performance, it is a simple matter of serving another model without that fine tuning in the enterprise API preferably only to the grok coding harness (or cursor, now). Coding performance for subscription plans don't move the needle in terms of revenue anyways and quality there doesn't matter as much.
> They are also buying 300-400 employees with proven record of training good coding models.
Are they? Their Composer 2.5 models is based on Kimi K2.5, it's not a bespoke model.
Achieving the improvements from K2.5 -> composer 2.5 with just post-training is actually more impressive. Though I believe their next model is trained from scratch.
> So SpaceX is buying 15 years of future cashflows, assuming no change in revenue. Which is bad assumption given past growth was gigantic.
I'd argue it's a bad assumption in the opposite direction. There's no moat. People can and will switch tooling and Cursor could easily be left with a steep decline in users.
ARR is best month times twelve, not revenue per year.
Is this what Cursor uses or a standard I'm unaware of?
I thought it was latest month?
Tbh I don't know if we can use traditional DCF calculations for things like this.
The main challenge is: If models get better, why would humans need a tool like cursor, when they have AI agents doing the coding for them?
But revenue is not really the informative quantity. If you sell gold you will have a huge revenue, but very little profit. I can be a trillion dollar company too if we exchange dollar bills for face value.
"training good coding models" many would say that is a highly debatable statement, and some would say that is just flat out not true. Cursor has not trained a frontier model from scratch, what they did was take an already made (non-frontier) model and further trained it on their user data about coding outcomes from its coding agent. So, a form of distillation and RL.
Revenue is not the same as cashflow.
It's $60B in SPCX stock, not in cash.
https://techcrunch.com/2026/06/16/spacex-to-acquire-cursor-f...
Guess Musk figured out that paying all cash to acquire something was a bad idea.
Sure - why use cash when you can use bits of paper instead?
I'd expect more of the same to come - good way to lock in some of this crazy SpaceX valuation by converting it into something with a bit more inherent worth.
Tesla next?
cash is bits of paper...
Not everywhere ... many countries have polymer banknotes for better security.
In any case not all bits of paper are equal. Monopoly money bills are not worth face value!
They are, its the exchange rate that's the killer.
Not even an IDE. A workflow built on top of an opensource editor.
Unsure if you're serious, but if you are, they wouldn't buy with cash, at least not the vast majority of it.
Cursor is massively overvalued. But so is SpaceX so it all evens out in the end.
> each share of Cursor’s common stock and each share of Cursor’s preferred stock outstanding immediately prior to the Effective Time of the Merger will be automatically converted into the right to receive shares of the Company’s Class A common stock based on an implied equity value of Cursor of $60.0 billion and the price of the Company’s Class A common stock equal to the volume-weighted average closing price thereof over the seven consecutive trading days
Current market cap is 2.66T which is pretty bonkers. Thats about intel, amd, and micron put together.
Its an all stonks deal.
Money laundry
its for the distribution, talent, proprietary data. spacex has the compute, and elon wants to win in AI.
Nope. They pay with the monopoly money, dilluting the shareholders.
90% of my dev workflows have moved to Devin cloud agents.
I don’t miss the days of fumbling around with my local repos across my multiple agent work trees or clones.
I just throw a task at Devin and I get a PR a few moments later.
Then it monitors the PR for any failing CI or review comments without me in the loop.
Now I can have 10+ Devin’s running at any given moment as I walk home from the coffee shop.
I cant believe people still use Devin. This is just filling the niche of vibe PR'ing? Not even vibe coding? Another level removed from the codebase than even claude code/codex are?
Yep, but the code is still reviewed/tested/validated before merging (so -vibe)
I'm genuinely excited for Cursor Composer 3. A Cursor model with Grok resources could compete with OpenAI and Anthropic.
Wasn't that already announced few weeks ago, only with deal going through depending on Cursor future stock price?
That was future option, now they are purchasing it.
They needed to raise money for the purchase, they just did it (raising from public market)
I love cursor. It's so frustrating that large companies are allowed to buy everything. Think of a world where github was still independent, cursor remained independent, heroku wasn't part of Salesforce. All great products that get eroded by the neglect of big tech.
The good thing is that cursor has no network effects. you could make your own (barely usable) cursor on top of vs in a weekend with the features you want.
I actually did that. Not on top of vs code because I wanted it on my phone. But now I have my own mobile IDE. It’s amazing
tbf, irobot being independent killed them. Sometimes it works well for these companies that would otherwise die. I strongly doubt cursor would survive against anthropic in 5-10 years.
I don’t think iRobot died because they were independent. I think it’s more a symptom of them stopping R&D investment because they were pushed to distribute cash.
I was (until this announcement/today) an annual subscriber just for the early days of Autocomplete goodness that I do not use anymore but would stick around and check from time to time their progress and always hoped we would eventually have a good UI for LLMs outside the terminal.
an IDE to look at one file at a time is not interface for LLMs - it was made for people, and while I guess you can sit in the Agents mode all day, to me thats a completely different byproduct and just "Terminal is yuck" kinda users, maybe someday we'll have a proper dedicated LLM UI but we are not there yet.
"In its IPO filing, the company had said Cursor's access to developers' data, including coding requests and design decisions, could help improve its AI models such as Grok."
This seems to be the key.. Data is expensive
What are the best open source IDE alternatives to Cursor? There was Continue for a bit but Cursor bought it. Is Visual Code by itself the only open source IDE atm? My main gripe with Visual Code is it doesn't make it easy to use open weight models or non Copilot model APIs. Continue helped but its now part of Cursor.
As for coding in the terminal, we have Opencode, Claude Code, and Codex etc. They are all open source but only Opencode can route to open weight models. But non of these are really an IDE like Visual Code or Cursor.
Are we missing a really good open source IDE to use open weight models? It seems like we are.
I think it's difficult to compete in this space because right now to build the "full IDE" you have to build an extremely capable harness (very hard) and a very good IDE (very hard). I plan to continue using Claude Code and built myself a small tool to verify what the agents actually do + move around the codebase quickly but it's a far cry from a full IDE: https://cotect.dev
Zed…
I don't know what's more bizarre, either it's SpaceX buying what is essentially a clone of VSCode with AI tools on top, or a VSCode fork being valued billions of dollars.
The value of code is approaching zero. SpaceX is buying training data and customer relationships.
So that accounts for about three quarters of the money SpaceX just raised then?
Not totally, because the deal is in stock. The cash SpaceX got is actual cash that can be deployed today. The stock will be in a lockout period and could be worth nothing or something whenever the lockout ends.
> about three quarters of the money SpaceX just raised then? While rest is paid to the debters...
Most probably it will be mix of stock/cash
It's all stock
I use cursor (through a work subscription), only the cli (https://cursor.com/cli), and mostly using Composer 2.5, but I freely change the model when the need arises.
Most comments here seem to think there is no command line client? I have never used the editor.
For my personal projects, I use a heavily modified pi. I also have access to a claude code account through work (bedrock), but I don't use it much. It always seems to be down.
The cursor cli (`agent`) is fine.
I find when working with just a single model (Codex / Claude Code), it tends to have more blind spots. So it's nicer to have different models review each others.
Also with Cursor you can make plans with more expensive models and execute using cheaper ones like Composer 2.5.
This is the best outcome for Cursuor. However, not sure if it's a good investment on SpaceX side though. I remember using Cursuor in early 2025, but after switching to claude code, I hardly used cursuor or even their agent cli tool. The competition is getting quite tough for all the AI tooling and I'm very curious to see how SpaceX is going to do with Cursuor
this is financial engineering to increase ARR temporarily, feels like next year Elon will dump lots of stocks
SpaceX valuation isn't based on ARR so juicing it isn't really relevant. And if he wanted to buy ARR for financial engineering reasons, he could buy a lot more than $4B with $60B
That’s two zeros too many.
Local inference seems to be catching up and Pi seems to be leading the pack for open weight harnesses. Bold move, SpaceX. I truly hope it doesn't work out for you.
How are these numbers even working out, I get free markets and all that, but Microsoft paid 2.5B for Minecraft, which was printing money at the time (seems they still lost on that deal). Now a rocket company is buying an editor company for 60B and everyone seems to think that makes sense.
I’m happy to be old man yelling at clouds here because I can’t for the life of me figure out these valuations and purchases.
All I know is, it's going to be fun when everyone wakes up sober and hungover realizing they've been sleeping on piles of tulips.
There's a lot of investment in AI for its potential.
An AI editor company might never make 60B itself, but it might help another AI company grow faster (relative to its competitors, who might also want to buy the AI editor company).
What else can an AI giant do with all that money?
Build in-house: they do, and there's only so fast they can hire/build.
Save? Yes, still do, but if they save it all, and let competitors buy Cursor, they lose.
Invest in other fields? Sure, but if they lose the AI race, that's all they'll be left with.
Tesla's IPO is a bet that if Musk has the right opportunity, he will do well. So he's given a big bucket of money, and needs a team that can deliver. So he buys Cursor.
The winners are Cursor. The losers are whoever is funding the AI companies that get outcompeted.
(Full disclosure: I don't know anything about Cursor, nor much about Tesla or its IPO.)
But surely 60B could buy you something better if you want to spend money on AI. The number seems completely arbitrary, would Cursor say no to 40B? I really don’t see Cursor as bringing them anything of actual value, it’s more of a bragging thing, but I can be very wrong.
Couple things. First, it was a all-stock deal, not cash, so it's closer to a merger than anything. Second, this deal was determined months ago, most likely at a equity stake defined then. That could mean that a 30B deal 3 months ago is now worth 60B.
> How are these numbers even working out
You must be new here.
Curser is the only possible play to make Grok AI some how useful to enterprise, where the vast majority of spend is / will be? Feels like it could only possibly worth this much to Musk. For me Anthropic are too far ahead - this will likely all be for zero shareholder value when Anthropic continues to pull ahead. Will be interesting to see if Musk switches off Collosus to impede Anthropic at some point.
For about 10 billions over the twitter price (inflation adjusted)
Mesmerizing....
I am still trying to find a replacement for cursor. But in the past 30D my automations and my own coding has consumed well above 10 billion tokens for less than 300$ with auto and composer 2.5; while building a fairly stable product with 20-30 daily active users. It feels like it’s too good to be true, because I’ve tried with Claude and codex and it just feels so much more expensive.
I'm excited to test out Zed's offering. I'm also going to look at some of the open source agent harnesses, even though I'll have to give up a fixed cost subscription. I suppose now is probably also a great time to kick the tires on self hosted models.
Composer 2 is actually good. I don’t mind when I run out of tokens on the mainstream models.
Everybody remember when Zuckerberg told in an Interview in 2024 that human data does not matter that much or more specifically "individual creators or publishers tend to overestimate the value of their specific content". Something along the line RL-Loops are more important.
Hard to square this with that acquisition which seems to be focused on Cursors vast amount of User Data.
At least for now.
Zuck is correct because the barrier to a large user base is not skill but market share which requires substantial effort, luck, timing, and patience.
Nobody can put a dent into Coca Cola because of their market. Better products exist but there is really no way to compete against $5 billion in marketing allowing them to maintain $50 billion.
Cursor is ~1MM users a day. $60k/per user is high but considering this is a stock buy, Space X "made" $300BN in the first day that is ~20% or one day of positive movement.
For Musk (with his baggage) to create or steer that user base would require a significant investment and time. Why not just slap some coupons from an initial bump to acquire the user base, user experience, and IP?
What an incredible outcome for the Cursor team. Hopefully the Cursor + xAI teams working together can produce a competitive frontier model.
Cursor was simply able to get early access to openai models and get an early lead doing things that are now obvious and done better by many others. Does anyone really want to use a crippled "enhanced" vscode to interact with a crippled version of codex or claude code?
This comment reeks of someone who has not used Cursor, at least in 2026.
Please point to a video or example of some way that Cursor is remotely state of the art at anything...
It's a great thing for consumers and businesses to have another competitive, American coding harness + frontier coding model duo. No one wants a crippled version of codex or claude code and surely SpaceX isn't paying $60 billion for that outcome.
> another competitive, American coding harness
Isn't the Cursor founder British?
So they innovated, created a product and found users? Sorry to break it to you but that’s how software companies are built.
I tried it and it was extremely broken and dramatically overhyped. Which I guess is what Elon finds appealing about it, as that seems to be his preferred type of business.
Do you think people are using Cursor to interact with Claude Code? What?
This should be an indication of how valuably xAI sees the training data that Cursor has accumulated, especially with its work on Composer 2.5. Last month, SpaceX and Cursor announced that they had been working on training a new model from scratch. Interested to see if this will put Cursor back into the zeitgeist.
Good timing because they are paying with SpaceX shares which are at a crazy high valuation right now (compared to just a few days ago)
What does cursor have? An ide and coding orchestration? They are using Claude or codex for llms, so they get acquired for their user base and tooling? Feels like a lot of money for that given Claude has the majority mindshare.
They have a giant pile of data from every customer that didn't disable data retention
I keep hearing this but they’ve been bleeding users since Claude Code came out so a bulk of their data is pre-Sonnet 4 and I’m not sure how the data from users prompting weaker models will help them now?
They did use that data to make Composer 2.5 which was decent but still a step back from GPT 5.5 and Opus 4.8. Though it’s really good at UI.
Not sure how much data they retained but it's possible they also have the _input_ tokens, and in the "early" days (how far everything moves in just a couple years...) more of those input tokens would have been non-AI-generated
I dont think so. I strongly prefer cursor over claude code and the composer model is basically unlimited and free also being super fast.
Of all the ai companies out there, anthropic and cursor are the two id invest in.
They technically have their own models as well, but they are based on other peoples models.
Most of this is just customers + staff/tech rather than models being acquired. Cursors actually got so much better in the last year. Their composer 2 model (a tweaked version of Kimi K2.5) is decent for day to day mundane tasks and the app can auto switch to more capable models when needed.
They have millions of conversations with code prompts and diffs and the user telling them the model is taking a wrong turn. This is quite valuable data that is probably already used to tweak Kimi code (composer 2). Moreover they have quite the enterprise client base that I suspect at least a portion of which will jump ship with the acquisition.
> What does cursor have?
users
People keep saying this but I have yet to find an integrated system that has good tab completion, cheap coding models and works well in an IDE. There are a number of options out there, none of them have captured much market share.
Zed seems to always suggest decent single-line comments for me, but I don't really use Zed for that that often.
Have tried it over time. I think it’s decent but I still found it to be not as great or quick as Cursor and the rest of their AI integration did not feel as compelling for my workflow.l but overall agree.
Isn't tab completion long dead?
Do you have a point? The whole package is compelling like I already said. Nobody has replicated it but everyone likes to ask what is the value but never share a compelling competing product. I am not Boris so I still write some code or make manual tweaks, autocomplete is nice.
Edit: I realize my question is maybe harsh but I think it’s valid for these drive by comments that drop a question like “isn’t tab completion dead?” Without any other substance it is a huge detractor to the comment quality of the site. At least add more substance or opinion.
All good, no offense from my side, apologies for the disturbance if it caused any. Generally would have just been curious if people still use tab completion much? Virtually all people I know have switched to Claude Code or Codex. I also read recently that Boris Cherny (CC creator) deleted his IDE. Of course that's marketing, but it shows the current state quite directly. So I'm wondering what your dev workflow looks like? Any why/when you use tab completion instead of coding agents.
No issues here but I think it helps to add a few thoughts to continue the dialogue.
I think it’s largely shaped by the type of work you are doing and I am not sold on the idea that we are at a point of never writing code. I leverage these tools a lot and there is still nuanced mistakes LLMs make that I have to intercept. I am still in the code at least for now. Other than my friends that work in startups I have not seen too much others delete their IDE and not have any touch on the code.
The have customers actually willing to use and pay for their service.
In a world where even Microsoft is needing to use AWS for capacity beyond Azure, xAI's utilization of their data centers has been so low that they are renting them out to competitors instead.
Nobody wants Grok. If you aren't using GPT-5 or Claude, you are probably using an open Chinese model like Qwen hosted by some provider.
I would expect Cursor to be forced to use the Grok Code models in short order. We'll see how people feel about "Mecha Hitler" writing their code.
Look at it the other way: What did Musk get last Friday? $75B from the SpaceX IPO.
It's almost like giving a toddler $100 in the toy department and seeing what happens next.
Customers, talent, training data, an increasingly competitive coding model and now a fuckton of compute.
So, you know. Couple of things.
Specially given that coding turns out to not be all that complicated, in the grand scheme of AI things: I don't think it's going take much more advances at the frontier before code writing will be as good as it need to be. At that point Composer (their model) catches up, what, 6 month later and they good.
Musk bought Cursor for its users (lots of devs still use it), as part of yet another attempt at catching up after building Grok and buying OpenAI failed.
Likely, Cursor becomes Grok Desktop or whatever, and eventually uses xAI's coding model if they can make a competitive one.
Yep, at that level one could end hunger in the world by 2030. But what does that worth compared to hold the destiny if the last shiny IDE.
https://www.ncbi.nlm.nih.gov/books/NBK599618/
It seems that each time there's a new tech cycle, another zero gets appended to all financial transactions.
Related:
> For the past few months, SpaceXAI has been jointly training a model with Cursor, which will be released in Cursor and Grok Build soon.
https://x.com/SpaceX/status/2066873915717136548
I wonder why IDEA didn't catch up with its own agent support and even their own models. It's not like agent harness is that hard to build, right? Or maybe they did, it's just that they haven't won the hearts and minds of the developers like Cursor did?
I think we should get used to it because that's what's going to keep happening again and again. First Twitter, then Cursor. When someone falls behind in the race for innovation, they usually buy the best product available and use it to get ahead of the competition.
Nonsense like this reminds me of the following quote from the 1999 Thomas Crown Affair movie:
“Have you figured out what you're going to say to your board when they realize you paid me thirty million more than others were offering?”
In the span of <20 years we’re talking about a sale price 3 orders of magnitude larger than a minor plot point of a hollywood movie.
Back in my day, a software company of national importance would be considered a big success if acquired for 60 million... Of course, at that point it would have to be profitable, and actually own its key IP...
That's interesting I was using Windsurf before and I really enjoyed it. Then it became part of Devon. I was less thrilled about that so I was looking at Cursor, but now it's also getting bought out. Any suggestions on what else is left? : )
My honest read is that, having everything — the data centers, the compute, the models (however misaligned they might be), the only thing xAI is missing is users. They don’t have any users because the only people who use Grok are essentially Elon’s fanboy club, and all they pretty much do with it is ask it to generate arguments to win their Twitter threads or nonconsensually unclothe people. Cursor gives xAI a captive audience of users; most sophisticated users don’t use it anymore, so anyone left is unlikely to be opinionated when models are shifted to Grok. Marriage made in heaven.
>the only people who use Grok are essentially Elon’s fanboy club
that's not true, Grok compelling feature is it's capabilities, performance and price. You only get comparable prices with GPT-5.4 mini or Gemini Flash. Also Grok Voice and images are pretty good.
I think at this point they don't want to be at the same level as Opus or GPT, they found their niche.
You generally see this sort of thing in the games industry when a publisher gives a developer a bunch of money to make a game but then the developer screws up and runs out of money. Publisher buys them to try to recoup their investment.
Is Cursor dying?
SpaceX is paying $60B in stock priced at 200x revenue. whether that's real money depends entirely on whether you think the stock is.
> The biggest focus of its business is the manufacture and launch of rockets with reusable parts.
Is it though ? Their TAM in their filing lists 85% as AI. $18.7 billion in REVENUE 2025 yet are spending more than 3x that for Cursor, and AI company.
Well that's a lot of money. They must see this as a distribution pipeline for Grok?
Fits right in: overvalued, over-promising, pure bubble hype.
I run Claude Code in cmux with Soonpatch and that's it. I tried looking at Cursor but honestly it wasn't providing any value and I prefer being 100% up to date with CC updates/interface
I’m happy for the Cursor team, but I sincerely don’t understand the valuation.
Most people I know who figured out how to use Claude Code or Codex directly get much more compute and a much better power-user experience. The difference is not even funny. And it’s not an IDE vs terminal because both claude and codex work just fine in vscode.
What annoyed me most was my own experience: last November I paid $20 for Cursor, burned through the quota in one day after just a few prompts, and didn’t find any clear way to even see quota, so i contacted support. Somehow they claimed I had used around $70 of compute and implied I should be grateful. But they calc price of all input tokens as if they were cashmiss tokens, which is obviously more expensive than real API usage and this was extremely dishonest on their part.
So from my perspective, Cursor or similar solutions often look like a middleman between the user and model provider. In theory, that should be hard to defend against using the frontier labs directly.
That is why it is mystery for me why so many people still pay for it over orders of magnitude cheaper claude/codex
Good for them. I stopped using it abruptly once I joined the claude cli + vim train but I am sure they will be happy to cash out. I think a lot of other devs stopped using it too so good timing.
I fail to understand, cursuor is bad at C++, really Bad. In fact of all tools and LLMs, cursor is worst, the best is kimi as per me. It know how to spot segfaults in long lines of code. Why would anyone pay for this hallucinating garbage, especially from a company which does mission critical C++
And, No It;s not a skill Issue. And No, my goal is not to make another vibe code music wrapper around ffmpeg, and No My goal is not make a program around some tcp wrapper. My goal is to fix and write new features in decade old code bases running across millions of devices.
I had stopped using Cursor since July 2025, because of its mess pricing policy.
Now I use codex & claude code & antigravity, and can not return to using Cursor.
There's no way in hell Cursor has $60bn of value This is insane.
$60B. Wow. Congrats to Anysphere. But $60B. That is just a ludicrous price.
Ugh, I'm already tired of seeing ads everywhere for Cursor about how you can build EVERYTHING and solve all problems using their agentic IDE, so now I have even more reasons to dislike SpaceX too.
I realized a while back that Elon Musk isn't Iron Man. His superhero (or supervillain depending on your view) persona is ZIRP Man, the master of riding successive credit expansion and speculative waves. It's sort of ironic that he at least pays lip service to some Austrian-style quasi-libertarian economic ideas, because the Federal Reserve created him.
Now he's surfing the AI wave. We are no longer technically in ZIRP but the delayed inflationary wave is now traveling through the economy and pumping everything. He knows the best way to soak up cheap money right now is slap AI on it.
I also had the thought the other day that him hitting $1T technical net worth might actually be a harbinger of a lot more future inflation. Inflation of this type hits assets before it hits things like prices and wages, and it hits assets with fast market cycles like stocks before it hits things like Real Estate. The blast wave starts at the top and moves down and out. So maybe Elon hitting $1T really means that in 20 years that'll be more like $100B inflation adjusted. Meanwhile a loaf of bread will be $20 and a starter home $4M.
But the fact that monetary inflation starts top-down is why low interest rates exacerbate inequality. The very richest and most leveraged can use the arbitrage gap to buy everything else before the inflation wave propagates. We've been in a low interest rate environment for about two decades, and you can see during that time how the super-rich with access to cheap money have fully detached from the rest of the economy.
In other words: the reaction to the 2008 financial crisis was to inject huge liquidity at the top, which created the new Gilded Age.
Ultimately it may be somewhat intentional. One way out of a sovereign debt crisis when you also have a sovereign currency is to inflate your way out, which basically is a huge tax on every non-domestic entity that owns your debt.
Elon Musk has exactly one talent, and it's being the greatest master of financial chicanery the world has ever seen.
To be honest and objective, I think he at least knows enough about engineering to hire people who know what they're doing, which is how he got here. It's not all chicanery.
Over time, though, I think he's drifted away from his original "make real things in the real world" focus and more toward "play money games" and "play political games."
It's sad. One common comic book supervillain arc is to start as a hero and become what you despise.
I really hope Musk doesn't ruin the product itself. I am not a fan of how he changed X at all. Id really like to see them stay on the current path, which has been brilliant so far
"Houston, we have a problem"
"One minute, let me ask my AI agent"
I had high hopes in Antigravity and Gemini, but they royally screwed up everything to such a level that the only worth plan is to use free plan for creating docs.
Will the $60B be newly minted SPCX shares? Does this effectively dilute investors by about 2-3% (minus some fraction representing the fair value of Cursor)?
Spcx is so overvalued that its value for the investors is whatever they wish it to be. If they wish it causes no dilution their wish becomes true. They already wished about 90% of the whole share price into existence.
Oh well, time to cancel my Cursor subscription. What a shame.
Good hopefully Grok Build gets better and they start to innovate in this area. Claude Code and ChatGPT need more competition. So that they also innovate more.
Been building a safety-first IDE alternative for 7 months, the acquisition makes the local/privacy angle even more relevant
What are the best open source IDE alternatives to Cursor? There was Continue for a bit but Cursor bought it. Is Visual Code by itself the only open source IDE atm? My main gripe with Visual Code is it doesn't make it easy to use open weight models or non Copilot model APIs. Continue helped but its now part of Cursor.
As for coding in the terminal, we have Opencode, Claude Code, and Codex etc. They are all open source but only Opencode can route to open weight models. But non of these are really an IDE like Visual Code or Cursor.
Are we missing a really good open source IDE to use open weight models? It seems like we are.
Those Mars bases are getting closer and closer.
Optimus robots are very close. Gradeschool vibes all around
Full self-living autonomous Mars base next year. Musk™ /s
Why would you accept payment in astronomically (pun?) overpriced stock? Maybe you have doubts about the true value of your own company.
They're doing what anyone would do in their position. I won't be surprised if they bought more companies while the stock price is that high.
When FB bought Instagram for $1B it seemed crazy too.
Yeah People thought the same about Microsoft buying Nokia which is more similar to this
Company flagging in one area trying to jump start its business there
Secure a cheap Cursor sub now before Elon jacks up the price. Supergroks ridiculous price is foreshadowing.
Cursor's stockholders better cash out quick.
Ah thanks for reminding me to cancel my subscription.
Good hopefully Grok Build gets better and they start to innovate in this area.
Claude Code and ChatGPT need some competition. So that they also innovate more.
If you are able to leverage claude code, neovim, and tmux effectively cursor is no longer needed in the mix, but to each his own
I wonder how much the founders are making here. They just started this a few years ago and now it has sold for 60B.
Why don't they just make their own in-house development environment? Cursor's codebase is maybe 5,000-50,000$ worth of actual code, even just $10M (compared to the $60B) could knock Cursor out of the park with a completely custom code editor, and even with a smaller budget they could fork VSCode. Maybe for the social value? I feel like a $10B advertising budget for a bespoke AI development environment is more than enough to become an actual competitor.
Distribution. Cursor has enterprise contracts, SpaceX/XAi want them.
Out of curiosity, anyone here still using cursor?
Yes it's my main AI thing.
It has access to all top models, great IDE integration and their AI based autocomplete is still unbeaten.
I have no desire to use a TUI, feels like a downgrade to me.
Yes, it’s my daily driver for building the saas I run full-time. I’m not happy about this news.
I like the ability to switch between any models, Composer 2.5 is really solid, I like having my agents coworking in the IDE with me, the plan mode is great, Cloud Agents are great, especially with slack, linear, web, etc integrations. I routinely tag an error report in slack and Cursor fires up a Composer 2.5 cloud agent that has readonly db access, access to error reporting, etc, and it can triage the issue, issue a PR, and tag me in slack.
The only thing I’ve felt like I’m missing out on is the subsidies of the CC/Codex subscriptions, but it seems like that is rapidly eroding anyway.
Curious, what does your SaaS do? even the general area is fine
It uses AI to replace a very niche human-powered workflow in a niche industry. That's all I'll say, but it's grown to about $30k MRR in the last year and is supporting my family, so the stakes feel pretty high to me.
I see, thanks for responding. Curious if this was an industry you were super familiar with?
Without revealing what your product is; how did you come across a good problem statement?
I've started on the bootstrapped train as well, also a senior engg.
I'd launched a pre AI software which grew to 5000 users and more and made me some money.
but post AI, I'm finding it hard to get into a non competitive industry. Like everything seems super captured already.
I actually acquired this as a little side project from some guy online. I hadn’t even really heard of the niche, but I knew enough to know that it had some potential, and I could also see that it could be used for an adjacent market segment that I was pretty familiar with, because my dad worked in that segment when I was growing up. And the price was cheap because it just had a few users and was pretty new. All but one of those users cancelled in the couple months after I bought it, because it turned out the product just didn’t do what it claimed to. But I didn’t really care about that, because 1) I knew it could deliver if built correctly, and 2) I could just tell there was demand for this and it wouldn’t be that hard to sell.
So I turned off the signup form and started rebuilding it and improving it, but I kept getting people emailing me and wanting to sign up. No marketing, nothing. I held them off for almost 18 months while I rebuilt it on the side of my day job, and that gave me a ton of confidence that I was on the right track.
Then about a year ago I got laid off. I was really close to relaunching it at that point, so I used all my severance to go all in on it.
Twelve months later it’s doing $30k MRR.
That's epic.
> used all my severance to go all in on it. Did you spend it on marketing btw? What did you find worked for your distribution?
Your story is super inspiring, so I had multiple questions, sorry haha. Want to be in the same places as you are.
Why are you not happy about this news? Didn’t their collaboration make Composer 2.5 possible?
Not to my knowledge? But even if that's the case, Cursor seemed like they were doing fine without SpaceX, and I'd like to avoid giving a single cent to Elon Musk. You can do as you wish.
I use CC/Codex/Cursor.
CC is mostly my default for large tasks / features (ex. Plan > execute plan ) Biggest gripe with Claude Code is that it is painfully slow relative to the other two.
Cursor for small stuff like bug fixes since it has a lot of models to choose from. I love the review/ diff / checkpoint features. It's planning feature is on par with CC. I'd probably use Cursor as primary driver if it had better cost efficiency. Next version or two of Composer may fill that gap in cost/quality/speed.
Codex isn't allowed at my work, but I use it for personal projects. It has the best balance of quality / cost / speed even if it's planner is poor and quite frankly the codex harness needs to catch up with the other two.
CC for quality / cost. Cursor for quality / speed. Codex for balance of the 3.
composer is competitive with around opus 4.5 in feeling?
largely lags behind opus4.7/gpt5.4, but is respectable, and generally outperforms the glm/qwen equivalents anecdotally despite benchmarks.
fails to follow instructions more often, and is less code critical, but performs okay if you can decompose the task to smaller problem spaces. i.e. only do manual review, only do typechecking, only do specific component. etc
https://artificialanalysis.ai/agents/coding-agents?coding-ag...
I agree, Composer 2.5 is really good. I use it for all kinds of small tasks, and really for any kind of first pass at debugging, answering questions about the codebase, pulling data for reports, etc. It’s fast, pretty accurate, and basically free.
Co-Pilot -> Cursor -> Claude Code.
I think my relationship with cursor was the shortest of all.
Cursor was really good for like 2-3 months. It felt like magic compared to Copilot.
Claude Code is like... I dunno, something better than magic because it actually exists.
Yes, I still use it, although less than I would otherwise.
Good: - Composer 2.5 is pretty decent for the quality / price ratio. - Easy to assign an issue to it in Linear (I know Linear just added this natively for linear agent, but it seems rubbish compared to Cursor) - Bugbot actually finds some useful issues (things Claude and Codex will miss) - Using @cursor in github usually works well, and better than @copilot. - Working with Python Monorepos with UV in their IDE. VSCode and Cursor work well here (Antigravity managed to screw it up somehow).
The Bad: - Usage/billing dashboards - These are are opaque and you can't attribute what actions map to what spend. - cursor won't follow PRs well like Claude Codes does. - Setting up environments is less good than Claude Code - Their IDE fork is woefully out of date, it'd be nice if it had more of the codeium fixes.
The Ugly: - Settings - Try to turn off bugbot, there's multiple places you have to do it. Good luck figuring them all out. - Support - they are polite, but gas light you and tell you it's your fault their product's settings are awful.
Linear employee here - if you have any specific feedback on our Claude/Codex integration, happy to hear it. Definitely a v1 so expect a number of fast follows up with some of the missing functionality like env customization, secrets, and code signing.
Cheers!
Cool.
I was speaking more on the linear agent vs the existing integrations. We love the linear guided reviews and issue tracking so have high hopes on getting a good DX here from y'all.
The claude integration - works as well as anthropic will let it work, since you can either automate it (anywhere from 'claude -p xxxx' or the api and ignore your subscription and pay by the token, or open a crapton of tabs with the terminal, or paste it into a bunch of sessions in their app. Which works more-or-less but it's cheaper than per-token costs.
The linear agent, doesn't seem to read the AGENTS.md file, follow along on a PR nor nor let you configure a sandbox (it told me this:
``` Note: I couldn't run ruff/pytest here (no uv/venv in the sandbox), so I verified syntax via AST parse only. The Postgres-backed tests will run in CI. ``` After I asked it to look at the PR check failures.
To be fair, claude code does it 70% of the time (the other 30% the sandbox is dead), and cursor about 10% of the time.
I intend to try it for design mode with Composer 2.5
Never did. Having been using Github Copilot since its launch (as autocomplete, they have a Vim plugin) and claude code for agentic coding.
I switched over to Claude Code. The products are essentially identical. This whole space has been commoditized. Paying $60B for this is idiotic.
> Paying $60B for this is idiotic.
Isn't that kind of Elon's thing?
Yes.
I was until today.
good enough for simple tasks.
SpaceX buys Cursor, and I just want autocomplete to stop suggesting closing braces I don't need.
I really dont get the point of cursor, I always find it subpar product too, but damn they are all got acquired congrats
Cursor is one of the lower performing agentic coding tools these days. Seems weird to buy a VSCode fork.
Interesting, Grok, for a flagship AI contender was rather poorly performing. I mean, not bad, but visibly less capable.
The grok fast models for coding were pretty adequate. Not good, but fast and good enough to be usable.
The writing was on the wall for cursor, this is a good deal for them to get bailed out and continue business.
Github Copilot is so much better than Cursor and a worldwide sales team. This acquisition has no chance.
If it takes $60B to respond to the prompt 'clone Cursor' I am not convinced it is worth $60B.
Congrats to the Cursor team. Unbelievable success. Can’t imagine how their live suddenly will change.
Further evidence that SpaceX is no longer a space company but a elaborate con to make Elon richer
Well that didn't take long.....
So that answers the question as to who's buying SpaceX stock…
Cursor is great but they're all going to cash out and leave SpaceX as soon as they can.
Sure why not. The points don’t matter and the game is made up at this point.
I’ve heard of space based microwave energy weapons. I haven’t looked at the physics personally but I talked to someone who did.
Something like a 10MW phased array to create a 1 cubic meter ball of plasma in the atmosphere. I wonder about the transmission losses… but what a weapon that would be!
I love cursor. The inspect tool in cursor is a game changer for UI work imo.
Didn’t they do that 2 months ago or are the re-announcing it because of the IPO
They reserved the option to buy it at this price, and are now exercising it
Damn. i see myself using Zed and CLI for codex and claude. But kudos to the team.
i thought they already did that!
they had the option, but hadn't executed on it... now it seems they have.
Will cursor launch a CLI tool like Claude/codex/opencode/pi ?
They already have. https://cursor.com/cli
Grok has its own CLI tool called Grok Build: https://x.ai/news/grok-build-cli
they have had one for a while now. https://cursor.com/cli
meanwhile all I need is a cursor (lowercase) and I can program and do all the actual software engineering I need. been true for many decades and likely will for many more
I think it's wild that Kimi is only valued at half as much - $30B.
ai is like the first technology with a conversational service manual inside it..
you should be foaming at the mouth to use claude or codex to make a custom harness, just for your own personal use with local models...
cursor acquisition and anthropic fable models banned from export by US government cant be coincidence. to keep powerful ai tools and model brains onside country
Well, grok build is an excellent product already IMHO
most likely it's for Cursor's data, not the IDE itself
I personally use CC and Zed -- works great. No need for a VSCode-based IDE. I've even dropped JetBrains (and I was a long-time user).
i don't see any point with SpaceX acquiring AI coding tool. it seems that they are just spending overflowing IPO money. am i missing something?
They can’t build something like Cursor with AI for less than $60B?
Ai is great. The bubble will burst. We will keep Ai just like we kept "the internet" after the dot-com bubble, but we still won't buy our pets online. I mean London has pretty good train connections and stations because a bunch of companies repeatedly tried to get rich. Most eventually failed, but we kept the rails. I just hope we get our computers back after this round of gambling.
I’m worried some of the people leading this race will try to entangle the institutional financial giants or even the government to ensure some sort of too big to fail scenario.
There’s an unprecedented amount of money at stake. And the admin has never been so openly and blatantly for sale.
Yes it’s a VSC fork, but it has an ARR of around 4bn
Whelp, I hope cursor was honest about not storing my data.
Not a chance
Can someone help me understand what this means to tesla and Grok?
I wouldn't be surprised to see Grok become some sort of agent that calls Composer/whatever the joint Cursor & X.ai model might be called.
The Infraggable Krunk
[dead]
Their Composer model (trained Kimi) is kinda amazing.
What are the alternatives to Graphite these days? Github's stacked PR support is still immature, AFAIK. I would love to see Linear move into this space and start offering both git hosting and stacked PR management.
My team uses git-spice [1], which is client-side only and works just as well as Graphite.
Aviator [2] is also good, and they have a hosted UI with merge queue support as well.
[1] https://github.com/abhinav/git-spice
[2] https://github.com/aviator-co/av
Nothing is a fantastic alternative. Graphite is an incredible amount of "extra" for a team that just communicates and knows how to use git..
I mean, I guess SpaceX has a lot of money to throw around. I just don't see how Cursor is worth even 10% of this number. There are so many similar tools out there - many plugins to VS Code (Kilo, Cline, Roo, etc), many CLI tools (claude code, opencode, plus everybody and his brother is putting out their own...).
I once again fear for my grandfathered-in free SuperMaven.
Interesting that they disclose this right after IPO.
Just like Allbirds, they're pivioting to AI.
I believe that OpenAI (I'll get to SpaceX in a second) has a huge valuation risk because:
1. It's a bet that OpenAI will "win" AI and have a significant moat; and
2. Future hardware improvements won't massively devalue OpenAI.
I believe open source models will win here, mainly because China won't allow otherwise. I also think that nobody is really talking about the hardware decpreciations coming in the next few years, which is going to be really important from a performance-per-Watt perspective. B100s aren't going to suck. But a theoretical T100 will get 30-80% more performance for the same energy input.
So, SpaceX. I've previously said that SpaceX would've been a significantly better company without xAI. SpaceX was used to rescue Elon and other "investros" from the financially disastrous Twitter purchase. Starlink, Starship (which is a risky program) and the Falcon 9 are a solid business. They're just not a $2 trillion business.
So I believe that the AI bubble contributes at least half of SpaceX's valuation and when and if that bubble bursts, at least half of SpaceX's value is at risk.
Google announced they're throwing billions to rent GPUs from SpaceX. That might sound good. It solves a short-term cash issue. But as another commenter put it, it makes SpaceX seem more like a Commercial REIT. After all, renting out your GPUs is literally the lowest-value thing you can do with them. You're not building a business. You're taking rent so someone else can build a business.
So buying Cursor and I'm sure any number of other AI startups in the coming year or two, seems aimed at kicking that AI can down the street.
So I view the Google-SpaceX as a red flag in the short-to-medium term. SpaceX simply can't seem to do anything valuable with all the compute they have. And I also have way more confidence in Anthropic (in particular), OpenAI and Gemini than I do in Grok.
> 2. Future hardware improvements won't massively devalue OpenAI.
I don't understand your point here. OpenAI rents all of its compute. It doesn't own datacenters. Any advances in hardware or optical computing benefit them because they can serve more customers. There is huge demand for GPT 5.5.
Use cash to buy GPUs and stock to buy companies.
That is very hinged
Money seems more imaginary than ever now.
Sadly, grok will always be shitty.
They will now own the whole pipeline, from computing to models and now the harness.
I honestly don't get why they feel the need to buy Cursor or why OpenAI wanted Windsurf. If it's data you're after, wouldn't it be so much more cheaper to just hire a dedicated team to fork VS code and integrate your own model and give it to the public for free with unlimited usage for a couple of months?
Vibe coded space shuttles baby! Lets GO!
Who is this Cursor person you speak of?
It's absurd. let's mark it down.
i hope they will integrate roblox for future generations of robot makers to be pre trained on games
Great! Just deleted my account.
Till the minute I clicked on this I couldn't decide if it was satire or not.
But also, for a few weeks periodicly I've been wondering what's going on with Cursor. Haven't thought about them at all, let alone used them, in quite some time.
They were a pretty big AI-native player. Seems clear we're well into the consolidation phase of this economic cycle.
We're in bizarro world.
Is this one of the situations where one of the engineers at SpaceX showed elon musk cursor and said "this is the future of programming" and that has been stuck in his head ever since? Otherwise (unless someone can prove me wrong) I don't quite follow buying out a company at the peak of their evaluation.
This truly shows how king making in venture capital is done, kids have the MIT pedigree but sometimes this is not enough for certain demographics, give them a ton of money to explore ideas and pivot, product is a vscode fork that sells subsidized AI, only possible with venture capital. Providing inference at unsustainable rate deemed as "product market fit". Product loses money until they exit.
VCs that say "I always knew the team was special", give me a break.
Oh, forgot I have it installed
yay -R cursor-bin
And gone
Glad I've jumped ship to Zed+ACP. I liked the idea of the new agent window view, but it's obviously slop and hasn't had a any polish or care put behind it given the number of bugs I keep seeing.
Zed is so much more stable and sleek and the agent view (threads) actually integrates nicely in a real editor. The side editor in the agent window was so much worse than the vscode base I expected, I have no idea how they dropped the ball so hard here.
Grok’s capabilities on Cursor’s data should be a step function there. Go X! Congrats Cursor, what a ride!
Definitely not a bubble.
It's been fun, bye
Take me with you
Our Beautiful Journey
Most of the comments here have never used cursor before and it really shows
At this point, money and rationality does not make sense to me, rather beyond my ability to understand. But I feel it is all about accounting and writing off eventually and a few profiting from it, not the retail investors for sure. Again I am just saying what I think and there could be no rational to my thought process.
waiting for the anouncement: cursor for grok heavy users.
SpaceX should rush to acquire as many companies as they can with stock. The market cap is absolute insanity (I know people keep saying this about new high scores in unrelated-to-reality valuations, but this one might just be the pinnacle), with zero rational basis, and they should try to make it real as rapidly as they can.
Next up, Anthropic.
Anthropic are already paying $15 b to space X for compute.
Buying depreciating nvidia hardware and renting it out to competitors isn't why SpaceX has a trillion dollar valuation.
true, can't hurt though.
> zero rational basis
Do you really think so? Like everyone who risks their and their clients' money here is just being irrational? Is this really a coherent view? Could it not be that someone knows something you don't, or does not have the biases you have?
Everything was in the S1 filing. There is no "secret knowledge".
The rational basis is entirely "I can sell the stock to somebody else for more money". Where in normal stock it would be "this company can make a profit that gives a return on this investment." This is a purely speculative play.
Why is it purely speculative for SpaceX and for nobody else? Is Musk just that good a con man, even though literally everyone hates him?
Why is something true for a company where it is true and not true for companies where it is not true?
SpaceX is worth more than Amazon.
Amazon has $750B of revenue and an enormous unfathomable moat across many fields, across most of the planet. It had a profit of $77B last year. And people consider Amazon overvalued, and a symptom of a serious bubble.
SpaceX has $18B revenue, and a consolidated loss of $4.9B. It has basically zero path to real profit, and it turns out that space launch actually isn't a lucrative industry, so much so that SpaceX had to create its own customer.
I mean, the biggest news about SpaceX has been the utter failure that xAI has been, reducing it to renting out all of the GPUs that Musk forced his other road-to-failure company, Tesla, to hand over to the failure that was xAI, that failed so badly it got hidden inside SpaceX. Good god. Somewhere in there the failure that was SolarCity got packaged in the giant scam.
Like seriously, the biggest win of the company is that it absorbed the husk of xAI that had a massive surplus of GPUs from when Elon tried to buy himself credibility in AI, and the market happened to make them worth more so that's their big win. Their biggest success is basically scalping GPUs.
>even though literally everyone hates him?'
Guy basically runs the US government, which has been reduced to a banana republic plutocracy. People invest in him because they know the system is so catastrophically corrupt that he'll always come up a winner, regardless of how enormous of failures he keeps generating.
I mean, this is such a transparent shell game scam that they're immediately talking about packaging up even more of Elon's scam businesses together. Just amazing stuff.
Don't worry, those super robots are coming any day now!
Yes, the US market is in the end games of a massive spiral -- a circle jerk of trillions of fake dollars moving in a rapidly accelerating circle -- and it will not turn out well. SpaceX is the moment when it is laid bare.
> Guy basically runs the US government, which has been reduced to a banana republic plutocracy.
Do you have any evidence of this? As far as I know, he basically never gets what he wants. He was against Trump's Big Beautiful Bill, it passed anyway. He wanted DOGE, it fizzled. He wants more Solar/Electric car related subsidies, Trump does not give a darn. He wants more H1Bs, Trump has been doing everything to frustrate H1Bs.
The one thing which Musk seems to have gotten was Jared Isaacman, but that was really difficult for him to get, and it took way longer than it should have.
Really difficult to see evidence that he runs the US government.
Point us at a rational verbalized or written argument for SpaceX's current valuation (and increasing)? Everything I've read says the valuation is too high and here is why, with x, y, and z reasons. Everyone I read who encourages buying seems to be ignoring arguments entirely and going on vibes.
I trust people who are investing their money more than analysts who get paid for writing what everyone wants to hear.
Interesting. I've not used Cursor in almost a year after using Codex/Claude.
Won't be surprised if Elon paid another hefty premium.
I just realized this whole game is just getting rich from other people's money and there might not be people left to buy those people's shares when the music stops.
It's literally a ponzi scheme.
Why after thier IPO?
[dead]
I can't stand Cursor. Every time I open it up I have 3 popups I don't use, that I then need to figure out how to close. Using it for notes is impossible, since the autocomplete just tries to fill in bullshit.
Awful what VC money did to it. Hope to never use it again, now that work stopped mandating it.
I quite liked Cursor. I even tried Claude Code and found myself wanting to go back to Cursor. Unfortunately, this completely kills it for me. I will not support Elon Musk or any of his shenanigans. He is already far richer than any person should be, but he also constantly tries to manipulate the government to benefit himself to the detriment of everyone else, whether that's DOGE, or the fast track to begging added to indexes on the stock market, or burying all the investigations into Tesla. I cannot in good conscious pay for a product when I know that he is profiting from it. So long Cursor, it was a good ride while it lasted.
On the flipside, this makes me MORE confident in Cursor.
But I already use Cursor w/ Starlink in the Cybertruck, so easy choice.
I also don't fall for propaganda so makes it easy to make practical, not emotional, decisions.
He's not evil, just a man with a vision, hopefully you realize that one day.
I don't care for his vision. I'm basing that on his words and actions, not based on anybody else's propaganda.
Another grab, I wonder there has to be something that someone pays $60B for, but my question is: I often struggle to understand why someone needs it. Is it extra plumbing and easement around code development or paying base?
"In its IPO filing, the company had said Cursor's access to developers' data, including coding requests and design decisions, could help improve its AI models such as Grok."
There it is.
This was a fantastically smart deal for both sides.
the IPO raised $85B and they just spent $60B on Cursor. If this was the intention it should have been in a disclosure
Edit: I see SpaceX did disclose
> With the option agreement, we have the right, but not obligation, to acquire Cursor at a predetermined price or pay a fee
> The consideration for the acquisition of Cursor, if any, after the closing of this offering would consist of shares of our Class A common stock based on an implied equity value of Cursor of $60.0 billion
"Collaboration with Cursor" page 12 of the SpaceX S-1 https://www.sec.gov/Archives/edgar/data/1181412/000162828026...
this is an all-stock transaction. No cash spent.
SPCX is exactly the "currency" that LLM companies should be valued in.
$60B cash? Too much.
$60B SPCX stock? Why so low.
didnt know about this
This is the most undeserved startup buyout in history. Literally anyone with a fork of VS Code could replicate Cursor in a very short period of time.
They should rename it to XCode. Oh wait…
They should rename it to CodeX. Oh wait…
SpaGrox
Cursor which is just a ux on other people's models and only has value as a UI that can be replaced by those models...
Jesus Christ, where do i buy cat bonds?
Is this Elon listening to Pieter Levels?
Pieter is so dumb. All he seems to do is post comparisons between the wonderful U.S. and dying EU that are completely wrong. If Elon is listening to Pieter, pray for Elon.
You do know Elon already believes those things?
Some of the talent at Cursor is second to none. E.g Less Wright, Sasha Rush, Stuart Sul.
Google paid 2.5B to bring Noam back into the fold in 2024 ¯\_(ツ)_/¯
Musk found a perfect market hack, buy a company at 10x their revenue and sell it in the stock market at 100x
Does that mean all of the co-founders would become billionaires? And they're, what, like 20 year olds?
And I'm here trying to get something to make a $1000 per month. What a world.
Just adding another "this is fucking stupid" comment for when this all burns down
incredible 60 B!
Aaaand now I'm using OpenCode instead, and trying out OpenCode Go.
but, why?
rip off.
Ponzi buys bubble for how much?
God fucking dammit I like cursor
meanwhile Mistral:
A fucking text editor?
Cursor is dogshit.
Good reminder for me to cancel my Cursor subscription. I don't support Elmo.
So there's yet another $10B+ company swallowed by the $T+ company.
"SpaceX told investors during the IPO process that it sees an addressable market for AI products worth $26 trillion, roughly equivalent to U.S. GDP."
This is unhinged.
I know it has become a meme by now, but IIRC the market for all foods (agriculture, processed food, etc.) on earth is around $10 trillion.
So according to SpaceX, the market for AI is 2.5 larger than all the food sold on this planet.
They're also saying that the AI market is worth roughly 10% of all global real estate.
This is true if you take the ai market as equal to the market for labor discounted to 5-10% penetration.
It’s not a totally unreasonable assertion, it’s the implication of the assertion that we are uncomfortable with. There is no reason for the models to stop their improvements in the near future.
> There is no reason for the models to stop their improvements in the near future.
Sure there is.
1. The cost of each new generation of training runs appears to be rapidly rising
2. The Trump admin just told the leading model to stop making it available to non-Americans, which in practice meant stop providing it at all
3. The factories to make the hardware are hitting bottlenecks, and while they've currently been navigated around, there's never a guarantee the next one will be
Currently I'm wondering at what point the direct impact on the US energy supply gives the US a taste of Baumol's cost disease as AI companies continue to outbid everyone else for electricity.
There are some counters to this, especially in electricity. We'll see massive expansions of wind and solar in the US because of this. Both the speed of install and low costs will guarantee it.
> We'll see massive expansions of wind and solar in the US because of this. Both the speed of install and low costs will guarantee it.
Implausible while Trump remains in office. He hates renewables, shuts them down even when doing so actively costs money.
Between AI hallucinated content and the politicisation of the numbers, I'm not sure how much AI compute capacity is being planned right now; would you accept a claim of 300 GW? It's a number I heard recently.
Given the capacity factor of PV, even China would have to think carefully before supplying that much PV over the next few years (300 GW avg ~= 3TW nameplate).
(Not sure about wind, wind's CF seems to vary between years).
Unless it's something on federal land businesses are pretty much ignoring Trump on renewables.
And 300GW power planned doesn't seem too far out of bounds, there are a huge number of 'planned' data centers all over the US.
World wide over 800GW of solar and wind was installed in 2025 and 2026 numbers should be over 1TW of renewables. How much of that will the US install itself is a much smaller percent, but as power prices increase the pace to profit off of it will quicken. I know China installed over 300GW themselves last year.
> Unless it's something on federal land businesses are pretty much ignoring Trump on renewables.
Except for all of the tariffs etc.
And that's without Trump seeming to be actively choosing winners based on favour to him, as with supporting Grok despite the data centre pollution in another thread and *possibly* (I don't wish to overstate my case) the ban on Fable.
> World wide over 800GW of solar and wind was installed in 2025 and 2026 numbers should be over 1TW of renewables. How much of that will the US install itself is a much smaller percent, but as power prices increase the pace to profit off of it will quicken. I know China installed over 300GW themselves last year.
This is why I wrote:
When the recent good news is that "the world installs 800 GW of PV" (TBH, I thought this was closer to the last 12 months of just PV than the sum PV+wind), that's the nameplate capacity, not the actual year-long-average output, which is about a tenth of that.The most recent PV capacity factor number on Wikipedia was 13%, which would make "800 GW" only 104 GW in real output; the figures I see for wind are that the CF is 25% (with much higher variability) but the nameplate capacity is lower, so they're pretty close as totals in real total currently installed output.
As long as Chinese companies keep pushing on, so must US companies too.
It would not surprise me at all if we suddenly start seeing top US AI companies lobby against Chinese models, or even the gov. making it illegal to use Chinese AI models.
But in this day and age, I just don't think it is possible. A distant third option would be that the big AI companies try to make hardware so expensive that people simply can't run their own models, while blocking access to foreign models.
> There is no reason for the models to stop their improvements in the near future.
You speak as if "improvements to models" is just function of time, and resources are infinite.
Models keep improving as long as there are resources to allow for larger and larger datacenters, if we hit a scientific breakthrough once LLM technology become the bottleneck, if the economy is infinite to allow infinite growth, and (geo)politics is not a thing to worry about. Or we discover ASI, machine improve themselves and we reach the technological singularity.
I know everybody is drinking the kool aid by the gallon, but can we maintain a little bit of objectivity?
yeah, it's funny how so many think the beginning of the S-curve is an exponential.
Granted, we don't know when the S-curve will inflect, but predicting too great an outcome is just as silly as discounting it altogether.
The s curve won’t inflect until it becomes difficult to allocate additional resources due to economic limitations. There is no sign that training a model on 10x the compute won’t lead to at least an equivalent improvement as the last order of magnitude increase.
If we define the Pareto frontier’s input in terms of a magic “compute equivalent unit”. We get a free order of magnitude from nvidia hardware improvements every 2-3 years. We get another order of magnitude from capital expenditure every 6-12 months. Kernel improvements to the models themselves likely yield an order of magnitude gain at some periodicity.
> the market for AI is 2.5 larger than all the food sold on this planet.
It just shows how much the automation has impacted agriculture and the food industry. Sure, there're rural farms that apply 200 yo technology. But e.g. the grain production and farming are incredibly efficient at scale. So, it's not that costly for as a humanity to feed 8 billion people (at a varying level, of course).
Remove 20% of AI supply, and the world goes on like nothing happened.
Remove 20% of food supply, and watch prices explode, global unrest, and famine take place.
Food is a solved problem. We can grow far more food than we need and we stop doing so simply because the low prices mean it's not economically viable.
In the places where famine remains a problem, it's due to political issues, not that we can't grow enough.
And growing all that food requires a tiny workforce compared to 400 years ago before the Agricultural Revolution. AI might extend such a massive reduction in labour requirements to many other industries.
> We can grow far more food than we need and we stop doing so simply because the low prices mean it's not economically viable.
Half. This depends on there being a reliable source of cheap fertiliser, which would be much more secure if not for the situations regarding Hormuz and Russia.
> Food is a solved problem.
Mmmmmh
> We can grow far more food than we need and we stop doing so simply because the low prices mean it's not economically viable.
So, it's not a solved problem. Last time I checked we have plenty of people in several parts of the world with difficulties to access the required level of food to be healthy.
> In the places where famine remains a problem, it's due to political issues, not that we can't grow enough.
The political issues are still there so I really don't think we can call that a solved problem.
That's what you misunderstand, that's why we're making the AI. Have the AI get rid of all the people then AI can grow all the crops it needs!
Technologically yes, but this is a vast oversimplification.
You need lots of money to be able to buy the tech you need to do so. And you can't exactly earn that from not using the tech, since foreign (or even local) competition will slaughter you on prices. And if you do make it, you're stuck with a low-margin race to the bottom on price.
So capitalism with market protections?
I don't think anyone is claiming AI and food have the same elasticity of demand, which is what this really talks to, but, after a claim the AI market is 26 trillion dollars... I wouldn't be surprised if someone did.
Valuation and elasticity of demand not related even if you ignore consumer surplus
Then I assume you'll happily pay $20 for a bottle of water, given how important it is?
do you spend most of your money on food?
I spend roughly 10% of my take-home pay on food.
I spend 0.2% on AI. Exactly one subscription.
do you spend most of your money on grok subscriptions ?
Most employers are spending a large fraction of SWE salaries on AI tokens right now.
It's not unthinkable that trend continues (even if it's rationalizing at the moment), and moves over into other fields as well.
I might at some point be spending more money on somebody elses claude subscriptions than on food
my point is that the amount of calories a person needs is limited, and the efficiency is non-decreasing, so the per capita spending has an upper bound
"ai" does not have such an upper bound
Ackshully, AI does have an upper bound in information theory, but since we're not anywhere close to writing data to the surface of a black hole I don't think it's a big issue yet.
Population growth isn't limited except through resources but AI also needs resources
We could remove 100% of world AI supply and humanity would not be worse off. It is still additive and in areas of generally indeterminate value except in hype.
Reasoning and RAG is amazing already and is a productivity gain but I'm yet to be convinced GenAI is anything but a slop machine.
#startflamingmenow
I half agree, but I'd still had software development to the list.
AI is useful as a search & information synthesis tool, and as a dev tool.
The problem is, when has a dev tool ever command such ridiculous valuations and investment in infrastructure?
The market is going to realize that yes, it's useful, but no, it's not over $1T useful.
> dev tool
do you expect people to quit research for some reason?
You can't food maxx a trillion calories a day to generate a multi million dollar bill. You can token maxx it though.
I think the issue is the reality that most life is worth a lot less (in US Freedom units) than some software running doing absolutely nothing truly valuable for anyone.
Nozick was right, the Utility Monster wins.
What precisely is the moat surrounding AI that SpaceX is using to justify this kind of spending spree? I don't how SpaceX and other AI companies will be able to keep the weights of their AI models private in the face of interest by virtually everyone in the world. It would be absolutely trivial for a nation state to walk into a data center using a state issued security certificate to seize a few of the physical servers running the cloud services of OpenAI / Grok / Claude. Copying the weights is trivial. Infiltrating a company with spies as new hire coders to gain access to source code is also trivial.
This is really starting to feel like the pets.com era again.
Because there's nowhere else for the money to go, the money must go to AI.
There are no growth opportunities in any other industry (except healthcare due to disastrous demographics), where else are people going to invest?
This is an insane take. Of course there are other areas that are growing or could grow if there was investment.
The problem is an absolute lack of vision on the part of those holding the capital.
Meeting the challenges of climate change could hold huge opportunities. Look at China’s massive expansion in renewables, look at the expansion of renewables in the US despite political headwinds.
Have some imagination, break out of your echo chamber. AI ain’t the only game to be played.
It's not lack of vision. It's that capital demands gambling. Everyone knows you could plough money into big projects in the US and double your money reliably. But the powers that be do not want that. They want to gamble, and try to become trillionaire #2.
To be clear I don’t see gambling to increase one’s wealth as a vision. That’s playing a game.
Vision is seeing a change that could be made. “I could be richer” is about as banal as a vision could get.
It's not insane. The GP is correctly describing a bubble economy.
The money chasing investments is orders of magnitude larger than the money people have on their pockets to spend. As a consequence, the only profitable thing to do is sell capital goods to make business and there is no profit on selling actually useful things.
China is in a different reality in large part because of their capital barriers that stop money from flowing in. Countries with bad reputation are also less affected.
What the GP gets wrong is that none of this makes AI a good business. Instead, it makes Nvidia a good business, but that's not news.
Looking at venture funding, it's definitely true. That doesn't mean other problems don't exist or aren't worth solving. But the concentration of (competing) capital and talent is insane.
The lack of places for investment money to go does not make the business profitable.
Not to be a jerk or anything, but that only matters for the guys and gals left holding the bag.
I don't think that will be Musk. He'll probably pull out significant resource from all this financial engineering relatively quickly. Probably via more financial engineering.
Cursor is a harness that can be used with all kinds of models. It's a much better harness than anyone else's and takes the company out of just playing the model game.
Exactly this. I think valuations and the AI market could get stirred up if:
- We get an open source Opus 4.8 equivalent and pair it with an open source coding agent
- Running this OS stack becomes cheaper than what frontier model providers charge (see OS model prices on OpenRouter vs. frontier lab prices)
- This happens across verticals (i.e. not just software)
The first “DeepSeek moment” didn’t do much damage back in the days, but I wouldn’t be surprised if a similar moment becomes a lasting, effective, cheaper alternative.
OpenCode exists, it is your "open source coding agent" that is practically on par with Claude Code and Copilot in terms of being able to do the 80% of things that most people actually use.
DeepSeek v4 Flash/Pro also exist, they are open weight and on par with Sonnet, just a bit below Opus. Again: practically useful and sufficient for 80% of things most people actually do. And most of the remaining 20% are benchmarks designed to push the limits, not productive work.
Using these already is way cheaper than your typical Claude API prices. What's still missing is a) mindshare - everyone still thinks "claude = coding" and everyone thinks he/she really needs the very best models because he/she is doing such incredibly complex stuff - and b) someone pushing such a stack as a convenient solution for corporations to easily dump their token money into, complete with user management, enrollment, monitoring, all that enterprisey stuff you need if you want to sell to, well, enterprise customers.
Need to give that a try then. What's your experience so far? Also in terms of cost/benefit?
My current hypothesis is that the $150 Claude Max subscription - of which I barely hit the limits anyways, even though I used it non-stop at work - still is very cost effective.
If the price of Claude Max increased significantly (say 2-3x), and my business would balk at paying the subscription, then I'd look for an alternative.
They are a distant third at best, at least in trading companies. If you look at Chinese and other likely national actors, they are probably further down.
The thing with dotComs is that they didn't have THIS level of unsustainable financing burn, and a tangible issue of token processing cost that has no magic wand coming with the current practical limits of Moore's law.
> I don't how SpaceX and other AI companies will be able to keep the weights of their AI models private
That is an interesting point. If there are higher concerns, copyright law is easily ignored, and only one person needs to get access to the data once.
SpaceX is good at building data centers in tough regulatory environments, in a way that other players have been unable to match
To be fair, food is the smallest bucket of my monthly expenses. And there are many people here on hacker news who pay more for their AI tokens than for their food.
How does argrar industry and tech industry compare as share of gdp in the US?
You are not in any way the average person.
Food is the 3rd largest expenditure in most households, after housing and transportation.
I'm sure people laughed in 2023 if someone said AI will reach $100B in revenue in 3 years. Yet here we are
Oh, that puts it 0.3% of the way there! And all it cost was increasing prices until every customer started talking about cutting it.
Past performance is not guarantee of future returns.
We'll figure out how to make it much much cheaper to produce the compute needs we have today with tokens. The question is will how many new use cases arise that need much more - clearly we aren't meeting needs (price stays high) but how much more do we need, 100x, 10x, 10^6 x?
Look at electricity, the world of 1900 could not create enough electricity or even conceive of how to add enough to meet 1950s needs. But we made it incredibly cheaper to produce, but also created a lot more, and boy do we have so much more use of electricity now. And it's not that expensive for a human to pay for their needs (not free, its not cheap for poor people but it's still gotten cheaper).
[dead]
“ai-x”? Elon? You know that you are not allowed to in this forum. It’s only for adults. Go back to Twitter and play with the other kids.
Food is worth a lot more than that. If the alternative was starvation, we would pay approximately all the money for food. By that metric food is worth more than $100T. The difference between $100T and $10T is called the consumer surplus, one of the largest benefits of a free market economy.
AI might eventually provide $26T worth of value, but if it captures anywhere close to that amount of revenue that'll indicate a failure of the free market economy. Competition and open source will have failed and the oligarchy has won.
(Either that, or inflation will have made $26T a relatively smaller number).
> They're also saying that the AI market is worth roughly 10% of all global real estate.
Why limit yourself to one planet? Space is infinite ;-)
I'm sure the finance market is much larger as well
>> AI products worth $26 trillion, roughly equivalent to U.S. GDP."
> This is unhinged.
The only way for Musk to become a quadrillionaire is hyperinflation. And a week later, we'll be quadrillionaires too!
Be honest now, it would take at least a few months for the rest of us.
Always fun to remember when calculating TAM - something like 85 - 90% of the world earns less than $1000 usd per month.
The math don’t math here, there literally aren’t enough people to afford this and businesses will go under the more people are displaced for gainful employment.
There have been a few recent stories about businesses finding themselves spending more on tokens than they were spending on the workers these AI Agents were supposed to have replaced.
Marks believe anything the con tells them as long as it's promising big money ROI.
You need to understand the definition of “total addressable market.” It’s a maximum theoretical number for the size of the market (not your company’s revenue) under ideal assumptions. A $26 trillion TAM is high but it’s not “unhinged.” For example, the logistics and transportation market is over $10 trillion and expected to double by 2035. Under ideal assumptions, if AI replaces everything from coders to lawyers, why is that “unhinged?”
Over what period of time?
By 2030? No way
By 2050? Maybe?
Obviously during an IPO you’re trying to make the bull case (unhinged or not). What does it look like in the best case scenario.
I also saw a quote from Musk saying that he expects SpaceX to hit $1 trillion in revenue by 2031. Given his track record of predicting performance I think it's safe to ignore such future looking statements from him or companies he controls.
Another way of putting this: global GDP is ~$132Trillion from what I gather.
So this is saying AI products will increase global GDP by about 20%.
The Federal Reserve says AI is contributing about 1% GDP growth per year to the US [0].
So maybe you can get to $13 trillion over a decade just from that. If you assume some acceleration, 20% isn't out of the question.
It is an extremely rosy projection, but if AI can replicate large fractions of the workforce, leaving those humans with the ability to work on other things, it doesn't seem unhinged when you think of it through this lens, just very optimistic - not Elon Musk level optimistic, just "everything goes according to plan and a bunch of things in the causal chain are all slightly on the higher end."
[0] https://www.stlouisfed.org/on-the-economy/2026/jan/tracking-...
> So this is saying AI products will increase global GDP by about 20%.
No business gets to capture 100% of the value it produces without physical coercion.
For infrastructure that requires high investment, it usually captures something around 5% of it. People tend to work really hard to replace or reduce any kind of infrastructure that gets near 10%. So we are talking about AIs increasing the global GDP by 200% at minimum, 400% more realistically.
Or in other words, bullshit number is bullshit.
> sees an addressable market for AI products
Well if you start adding AI powered to "everything" then it is possible.
Soon you'll have AI face cream and AI donuts.
I call it forward thinking: they assume massive inflation due to income taxes breaking away.
> This is unhinged.
Just like the investors :D
Where is that quote from?
I can’t see it in the article when reading on my phone?
Dunno the actual number, but one thing I'm certain is that it must be closer to $0 than $26 trillion on a number scale.
yea, totally nuts.
clearly it's more like $540.2 quintillion at this point
They are pricing in inflation along with the inevitable money printing that will happen.
In a sensible world, this would be considered lying to investors and be prosecuted.
[dead]
With SPCX shares never going down in price, SpaceX can acquire all companies in the US in exchange for its stock, so SpaceX itself is worth at least as much as the US GDP! (/s)
[dead]
[dead]
[flagged]
[dead]
[flagged]
[flagged]
[dead]
[dead]
[dead]
[dead]
[dead]
[dead]
[dead]
Well the posts here will be rational and well-informed, I'm sure.
/s
[dead]
[dead]
[flagged]
[flagged]
[flagged]
It won't be expedited into S&P.
[flagged]
Congrats to the founders, arguably the first true AI-wrapper billionaires? 0 to multigenerational wealth in 4 years is impressive. It is crazy how much more wealth per user that can be created in the AI-space, compared to other products.
It is crazy how much more wealth per user that can be created in the tulip space, compared to other products.
Surprising how tech people on a tech forum are some of the biggest Luddites. Maybe it’s because the creative destruction is coming to your industry this time?
Crazy how AI has become a sort of religion beyond question for some people. AGI is like the big nerd rapture they’re all waiting for, any day now.
The irony is that once AGI arrives, tools like cursor would be worthless.
In fact, if AGI arrives, and it is possible to run such a model / models locally, the whole idea of commercial models would be a bit dead, yes?
Is he bailing out an investor he's connected to?
Cursor has a very cluttered UI. Very unintuitive.
[flagged]
Ok, but please don't post unsubstantive comments to Hacker News.
Not the entire industry, only the American part. Chinese companies seem healthier.
You can only say that because you know nothing about the Chinese ones
Yup
Just cancelled my subscription.
I've been using the Pi agent with Deepseek for some days.. and I'm more than happy with that.
This will ultimately be good adding another competitor into the mix with a very strong coding dataset + enough compute to make aggressively top tier models.
This is the card spaceX needs to play to be able to get composer / grok to complete w gpt and Claude
This will be a net positive for our entire ecosystem from a progress and options perspective.
Expensive price but great for cursor shareholders and plenty of demand of spaceX stock at this crazy high valuation.
I don't own spaceX stock at this price
I think Starlink will vastly out perform projection I think datacenters will under perform I wonder how many nvidia chips spaceX locked in for next 2 years and I think that numbers is actually the most important number
Cheers