There are only two successful ad business forms in the digital world, attention or intent. Meta is built on attention and Google and others like amazon are built on intent.
Everything else is optimizing the targetting data in some kind of behavioral way to get better intent data or to reach the right users.
I have no idea where something like Chatgpt stands on that axis, it has actually very little attention (in hours per day) for most people and I am not sure that it has enough intent signals.
> Meta is built on attention and Google and others like amazon are built on intent
I'm having trouble seeing the difference. They both capitalise on attention, i.e. eyeballs. And they both try to predict what you're likely to click on so they can show you the most-relevant ads.
I get that from the product perspective. I'm just not seeing it from the ad side. Attention is a prerequisite to serving an ad. Correctly predicting intent is how you maximise its value.
Parent is presumably making the case (and I'm assuming this extends to their respective ad pricing models) that it's revenue equivalent but profitable either way.
Either you have intent but little attention (Google et al.) and so can charge a decent amount.
Or you have lots of attention but little intent (Facebook et al.) and so can charge a decent amount.
You can't charge a decent amount if you have little intent and little attention.
OpenAI presumably has intent, but hasn't done all the Google-esque adtech sausage making to expose that to ad space buyers.
> When you use Facebook you get ads for things you aren't looking for.
When you use Facebook you get ads for things you are looking for in a broad sense, they just happen to figure out what you are looking for via fingerprinting and spying on all your web and app activity (to the maximum extent allowed by platforms and regulators).
Google, of course, also does this in addition to looking at your currently active web search.
Sure, they retarget ads they think you might want to see, but the difference is when using Google you are actively looking for something, not passively.
Sure but I still don't think there is much meaningful difference.
Google looks at both your current search/browse activity and gathered data (the gathered data being especially important for all the very many Google-served Ads off of google.com)
Facebook also looks at your current search/browse activity and gathered data.
The only real difference is Facebook search is universally entirely worthless so far less of that signal gets used relative to more passive snooping.
Emarketer’s data finds that standalone chatbots like ChatGPT,
Microsoft Copilot app, Google AI Mode, and Amazon Alexa for
Shopping (formerly Rufus) in U.S, will generate less than $1
billion in ad revenue this year, and just $5.41 billion by 2030.
How would that be possible? In the past I used Google maybe 10 times a day with a short query. From which Google had to guess my intent. Now I babble with Gemini all day about everything. And Gemini can ask questions what exactly I mean. Why wouldn't Alphabet be able to generate more revenue from this than from search? And Google's ad revenue from search is over $100B per year.
Just because there are no ads now does not mean there never will be. Google search was run without ads for the first years too.
> How would that be possible? In the past I used Google maybe 10 times a day with a short query. From which Google had to guess my intent. Now I babble with Gemini all day about everything. And Gemini can ask questions what exactly I mean. Why wouldn't Alphabet be able to generate more revenue from this than from search?
When you used google 10x a day, you saw, at a minimum, 20 ads/day. How many ads are you seeing when you babble all day long with a chatbot?
> When you used google 10x a day, you saw, at a minimum, 20 ads/day. How many ads are you seeing when you babble all day long with a chatbot?
There is also way more competition between chatbots than there was between search engines after Google started leaning into ads. Which strengthens the Khan-Esayas hypothesisis "that data privacy constitutes a key parameter of non-price competition in the market for" consumer tech [1][2].
why would they be able to generate more revenue? it isn't clear to me. Maybe ads will be targeted better, maybe they wont. Either way, major companies marketing budgets won't be going up, only shuffling around?
So its not really much to do with the technology or how good it is, but the stories their salespeople can tell. And I guess the story isnt good/convincing enough
The same kind of speculation questioned whether Facebook would ever have a significant business. Before that, the same was asked about Google's business.
If GPT can maintain or grow usage, the ad dollars will be there given the enormous scale. There is a hundred billion dollars plus worth of advertising waiting in the LLM space. It would be surprising if Facebook doesn't contract for example, losing ground to LLMs on advertising over the coming decade.
They have to build out all the backend ad tech crap that google and fb have built over decades. As in the campaign tools, scheduling, targeting, analytics, auctions, exchanges etc. Advertisers and publishers have already deeply integtated into that eco system. Also the google and fb system has created a content generating army that is incentivized to fight tooth and nail for eyeballs. OpenAI doesnt reduce the activities of the eyeball capture army, it just freaks them out and pushes them into overdrive.
WhatsApp and Messenger have billions of users but don’t make much ad revenue. I’m sure there exists a way to show ads in ChatGPT that will be figured out eventually but so far nobody’s figured out how to monetize chat as a medium
Also it’s basically free for Google to show you a sponsored result but embedding one in a ChatGPT response actually costs money (assuming they’re part of the generated response).
Lastly I will bet you one Stargate datacenter that Meta has thought about LLM-based advertising, and if there’s any low hanging fruit there it’s already been tried
> WhatsApp and Messenger have billions of users but don’t make much ad revenue.
Not directly, but I bet that connection graph is worth millions. Same with the Giphy keyboard purchase. There's a whole unseen economy of data being valuable for reasons that aren't widely discussed.
Facebook was in financial loss until it got really serious about pushing ads and outrage. People who speculated that Facebook is not earning much on ads while Facebook was not earning much on ads were correct. Same with Googles business.
And also, very relatedly, both companies enshittified their products great deal in the process. Google search used to be better for average user, Facebook used to be better for average user.
Genimi can start pushing ads everywhere, can start using dark patterns to better manipulate us and one they will. They will start earning more money at that point.
Ignoring whether or not OpenAI can reach this, do we think a massive expansion of ad revenue is fruit of the poisonous tree for GPTs? Does anyone want to use a tool that is capable of disguising an ad as a geniune recommendation at anytime? I suppose if the platforms become super entrenched and we boil slowly then they'll get away with it
> Does anyone want to use a tool that is capable of disguising an ad as a geniune recommendation at anytime?
It depends on the economics and usefulness of the tool. Google Search got away with it because it was useful AND trading zero cost of usage for ads was worth it for most.
Whether LLMs can get away with it seems very unclear now as they don't seem as popular as search engines.
LLMs don't seem as popular as search engines? We are all living in our own bubbles, so I'll tell you about mine. In mine, LLMs have replaced search engines. Not just the query itself, but how it's used. I don't use Google to find a site that tells me what to do after I've spilled some red wine kny the carpet, I use ChatGPT to tell me directly what to do about it. That's foundational.
Founding a start-up with VC money, dumping the bag into someone else's lap, and then riding off into the sunset is a very different skill set from actually growing and running a company that is no longer early stage.
Look at Zuckerberg. He may have been good as a founder, but if he didn't have super voting shares, and if Meta's cash flow from their ad business wouldn't mask how poor his strategic decision making has been, he would have been pushed out as CEO years ago.
> I reckon they'll get government to block chinese models just like the US car industry did with EVs.
This is more or less already the case. Not that they're blocked, but... it may limit who you can work with and what insurance companies will cover you.
tldr is: They either get a successful IPO to stave off bankruptcy for a couple more months, or they're going to be bankrupt by the beginning of next year.
Or they stay private and raise more capital. Until they have a failed round people predicting bankruptcy are getting way ahead of how this would actually play out. Some of us are old enough to remember the "Amazon can never make a profit and will go bankrupt" predictions of 25 years ago.
The early years Amazon wasn't profitable by choice. They could have stopped that at any time and even did demonstrate it by having a single quarter with $1 profit or some such.
Anthropic and OpenAI have no choice but to go public if they want to avoid bankruptcy. Venture capital firms are struggling to raise more capital, the bond market is so saturated that the borrowing costs are getting too high and big tech is also at the limit of how much they can invest, all while AI companies' costs are going through the roof. Retail investors is the last market they haven't tapped into and to do that, they have to go public. There's just no way around it.
The tech industry is currently spending more on AI infrastructure every single year than the United States spent during the absolute peak annual years of the post-9/11 wars.
That said, $2T was spent during GWOT with another $8T in veteran care, DHS, interest in debt.
Current AI spend this year is expected to be $2.59T (chips, infra, etc)
There's a very low chance of OpenAI doing an IPO this year: https://polymarket.com/event/openai-ipo-by. So the assumption that they "have to do an IPO" already seems questionable. The rest of the analysis also feels pretty hand-wavy. I'm not saying OpenAI is in a strong financial position, but this article doesn't make a solid enough case for why it's supposedly in such dire straits.
"It means that OpenAI could soon rack up losses that exceed its asset value. In other words, they are on course to slam into bankruptcy this year."
...that's not how bankruptcy happens. What is this guy's background?
Assets on a balance sheet are held at book value. You can absolutely run GAAP losses that exceed net assets without running into bankruptcy, particularly if you're granting (and having employees exercise) options.
The critical measures are cash in and out and debt-like obligations. None of those metrics point to OpenAI going bankrupt this year unless they do something really fucking creative. (Which, to be clear, is Altman's M.O.)
You're right of course, but in this case the statement doesn't look that false. The source article [0] states that half of the assets is cash, so "assets" is defined a bit more freely here.
If we assume a 2026 revenue of 23B (Q1 2026 revenue x4) and costs of 70B (2025 costs * (2026 revenue / 2025 revenue)), even those 25B in cash reserves won't mean much. And this napkin math even ignores their debt obligations. So either they raise money in the range of 50-75B this year (actual money, not datacenter vouchers), or ...
> costs of 70B (2025 costs (2026 revenue / 2025 revenue)), even those 25B in cash reserves won't mean much*
If all those costs are cash costs, sure. When OpenAI gets non-cash investment, and it "burns" that investment on compute, that should be counted as a GAAP expense. Yet it doesn't touch cash. (It would destroy a unit of compute asset. But again, book versus market value can mess with how that works intuitively.)
Similarly, if you're granting lots and lots of options they're going to generate lots and lots of compensation expenses (and thus losses) as they vest. These aren't cash expenses, however.
A simulacrum, simultaneously more detailed than the ones built by Meta and Instagram and Tik Tok, but even more shallow, without even the illusion of human connection.
An endless forest of mirrors reflecting no one but the user.
True, but the current state of advertising is, charitably, kinda goofy.
I'm not surprised the iteration we're seeing now is perceived as failing: on the few remaining screens where ads are still present for me, they're between highly irrelevant to flat out repulsive. There are a few brands I will never touch with a long stick, purely as a result of their disturbing, disgusting ads -- again, charitably, a negative-sum game.
As someone who doesnt really see ads either, we are in the minority. The vast majority of people are bombarded with ads from their TVs, streaming platforms, and unadblocked browsers. Try using a relatives iPhone or something its mad.
There are only two successful ad business forms in the digital world, attention or intent. Meta is built on attention and Google and others like amazon are built on intent.
Everything else is optimizing the targetting data in some kind of behavioral way to get better intent data or to reach the right users.
I have no idea where something like Chatgpt stands on that axis, it has actually very little attention (in hours per day) for most people and I am not sure that it has enough intent signals.
LLMs are great for product research, as in "I want to buy X, what brands are there and what are the pros/cons and what do users report".
Lots of potential to game the results there. If that is enough to sustain the business is another question.
They are only great right now because, unlike web search, results are not yet sufficiently gamed and monetized.
There was a time, in an age long gone, when a cursory web search was great for product research too.
Wouldn't the idea be to leverage inattention: people delegating their purchasing decisions to the chatbot without realizing that it's on the take?
> Meta is built on attention and Google and others like amazon are built on intent
I'm having trouble seeing the difference. They both capitalise on attention, i.e. eyeballs. And they both try to predict what you're likely to click on so they can show you the most-relevant ads.
When you use Google you are generally actively looking for something. When you use Facebook you get ads for things you aren't looking for.
I get that from the product perspective. I'm just not seeing it from the ad side. Attention is a prerequisite to serving an ad. Correctly predicting intent is how you maximise its value.
Parent is presumably making the case (and I'm assuming this extends to their respective ad pricing models) that it's revenue equivalent but profitable either way.
Either you have intent but little attention (Google et al.) and so can charge a decent amount.
Or you have lots of attention but little intent (Facebook et al.) and so can charge a decent amount.
You can't charge a decent amount if you have little intent and little attention.
OpenAI presumably has intent, but hasn't done all the Google-esque adtech sausage making to expose that to ad space buyers.
> When you use Facebook you get ads for things you aren't looking for.
When you use Facebook you get ads for things you are looking for in a broad sense, they just happen to figure out what you are looking for via fingerprinting and spying on all your web and app activity (to the maximum extent allowed by platforms and regulators).
Google, of course, also does this in addition to looking at your currently active web search.
Sure, they retarget ads they think you might want to see, but the difference is when using Google you are actively looking for something, not passively.
Sure but I still don't think there is much meaningful difference.
Google looks at both your current search/browse activity and gathered data (the gathered data being especially important for all the very many Google-served Ads off of google.com)
Facebook also looks at your current search/browse activity and gathered data.
The only real difference is Facebook search is universally entirely worthless so far less of that signal gets used relative to more passive snooping.
Just because there are no ads now does not mean there never will be. Google search was run without ads for the first years too.
> How would that be possible? In the past I used Google maybe 10 times a day with a short query. From which Google had to guess my intent. Now I babble with Gemini all day about everything. And Gemini can ask questions what exactly I mean. Why wouldn't Alphabet be able to generate more revenue from this than from search?
When you used google 10x a day, you saw, at a minimum, 20 ads/day. How many ads are you seeing when you babble all day long with a chatbot?
> When you used google 10x a day, you saw, at a minimum, 20 ads/day. How many ads are you seeing when you babble all day long with a chatbot?
There is also way more competition between chatbots than there was between search engines after Google started leaning into ads. Which strengthens the Khan-Esayas hypothesisis "that data privacy constitutes a key parameter of non-price competition in the market for" consumer tech [1][2].
[1] https://yalelawjournal.org/pdf/e.710.Khan.805_zuvfyyeh.pdf
[2] https://www.researchgate.net/publication/327111419_Privacy_a...
> How many ads are you seeing when you babble all day long with a chatbot?
You wouldn't see any of them, because you'd not be aware of the invisible bidding war over which products the bot recommends.
> When you used google 10x a day, you saw, at a minimum, 20 ads/day
That's assuming a user with no ad-blocker which is an ever diminishing number of users
We are talking about income from ads. It comes from people who see ads.
> Google's ad revenue from search is over $100B per year.
Google serves ads on search, maps, YouTube, android apps, amd all websites using Adsense or video ads
why would they be able to generate more revenue? it isn't clear to me. Maybe ads will be targeted better, maybe they wont. Either way, major companies marketing budgets won't be going up, only shuffling around?
So its not really much to do with the technology or how good it is, but the stories their salespeople can tell. And I guess the story isnt good/convincing enough
> And Google's ad revenue from search is over $100B per year.
From search or from adsense? That's very different. People place adsense banners on their websites which has nothing to do with search.
Because writing a headline with 90% miss in it will get more clicks than actually thinking through business strategy and realistic projections.
The same kind of speculation questioned whether Facebook would ever have a significant business. Before that, the same was asked about Google's business.
If GPT can maintain or grow usage, the ad dollars will be there given the enormous scale. There is a hundred billion dollars plus worth of advertising waiting in the LLM space. It would be surprising if Facebook doesn't contract for example, losing ground to LLMs on advertising over the coming decade.
They have to build out all the backend ad tech crap that google and fb have built over decades. As in the campaign tools, scheduling, targeting, analytics, auctions, exchanges etc. Advertisers and publishers have already deeply integtated into that eco system. Also the google and fb system has created a content generating army that is incentivized to fight tooth and nail for eyeballs. OpenAI doesnt reduce the activities of the eyeball capture army, it just freaks them out and pushes them into overdrive.
WhatsApp and Messenger have billions of users but don’t make much ad revenue. I’m sure there exists a way to show ads in ChatGPT that will be figured out eventually but so far nobody’s figured out how to monetize chat as a medium
Also it’s basically free for Google to show you a sponsored result but embedding one in a ChatGPT response actually costs money (assuming they’re part of the generated response).
Lastly I will bet you one Stargate datacenter that Meta has thought about LLM-based advertising, and if there’s any low hanging fruit there it’s already been tried
> WhatsApp and Messenger have billions of users but don’t make much ad revenue.
Not directly, but I bet that connection graph is worth millions. Same with the Giphy keyboard purchase. There's a whole unseen economy of data being valuable for reasons that aren't widely discussed.
Facebook was in financial loss until it got really serious about pushing ads and outrage. People who speculated that Facebook is not earning much on ads while Facebook was not earning much on ads were correct. Same with Googles business.
And also, very relatedly, both companies enshittified their products great deal in the process. Google search used to be better for average user, Facebook used to be better for average user.
Genimi can start pushing ads everywhere, can start using dark patterns to better manipulate us and one they will. They will start earning more money at that point.
Ignoring whether or not OpenAI can reach this, do we think a massive expansion of ad revenue is fruit of the poisonous tree for GPTs? Does anyone want to use a tool that is capable of disguising an ad as a geniune recommendation at anytime? I suppose if the platforms become super entrenched and we boil slowly then they'll get away with it
> Does anyone want to use a tool that is capable of disguising an ad as a geniune recommendation at anytime?
It depends on the economics and usefulness of the tool. Google Search got away with it because it was useful AND trading zero cost of usage for ads was worth it for most.
Whether LLMs can get away with it seems very unclear now as they don't seem as popular as search engines.
> as they don't seem as popular as search engines.
Anecdotal evidence that I observe is that people stopped using google and only ask LLMs nowadays for their everyday queries.
LLMs don't seem as popular as search engines? We are all living in our own bubbles, so I'll tell you about mine. In mine, LLMs have replaced search engines. Not just the query itself, but how it's used. I don't use Google to find a site that tells me what to do after I've spilled some red wine kny the carpet, I use ChatGPT to tell me directly what to do about it. That's foundational.
Why does it have to be ad revenue though? A forecast of ad revenue alone being below expectations really doesn't say much about a company like openAi
Does anyone at OpenAI know how to run a business?
Sam Altman worked at the same incubator that runs this website.
Founding a start-up with VC money, dumping the bag into someone else's lap, and then riding off into the sunset is a very different skill set from actually growing and running a company that is no longer early stage.
Look at Zuckerberg. He may have been good as a founder, but if he didn't have super voting shares, and if Meta's cash flow from their ad business wouldn't mask how poor his strategic decision making has been, he would have been pushed out as CEO years ago.
What’s the business? If it’s anything other than raising equity then you didn’t answer the question.
But how many companies did he actually launched? lol
The only actual product was "Loopt", some bs location sharing app that nobody used or ever heard about.
They should poach Neal Mohan.
Wouldn't be surprised if there is utter panic behind the scenes.
They somehow need to corner many billions of business meanwhile chinese labs reckon they'll have fable class models by end of the year. [0]
That does not leave a lot of room for mistakes. I reckon they'll get government to block chinese models just like the US car industry did with EVs.
https://www.tomshardware.com/tech-industry/artificial-intell...
> I reckon they'll get government to block chinese models just like the US car industry did with EVs.
This is more or less already the case. Not that they're blocked, but... it may limit who you can work with and what insurance companies will cover you.
So, how do their overall financials look these days?
They can probably raise more capital, and stave off the inevitable bankruptcy for just a bit longer. Wouldn't be surprised if US Gov buys a stake too.
Will Locket wrote about it last month: https://www.planetearthandbeyond.co/p/openai-is-in-a-far-wor...
tldr is: They either get a successful IPO to stave off bankruptcy for a couple more months, or they're going to be bankrupt by the beginning of next year.
Or they stay private and raise more capital. Until they have a failed round people predicting bankruptcy are getting way ahead of how this would actually play out. Some of us are old enough to remember the "Amazon can never make a profit and will go bankrupt" predictions of 25 years ago.
The early years Amazon wasn't profitable by choice. They could have stopped that at any time and even did demonstrate it by having a single quarter with $1 profit or some such.
Anthropic and OpenAI have no choice but to go public if they want to avoid bankruptcy. Venture capital firms are struggling to raise more capital, the bond market is so saturated that the borrowing costs are getting too high and big tech is also at the limit of how much they can invest, all while AI companies' costs are going through the roof. Retail investors is the last market they haven't tapped into and to do that, they have to go public. There's just no way around it.
The scale current of money-burning is just wild. I'm not sure you can compare it to anything else this century.
The tech industry is currently spending more on AI infrastructure every single year than the United States spent during the absolute peak annual years of the post-9/11 wars.
That said, $2T was spent during GWOT with another $8T in veteran care, DHS, interest in debt.
Current AI spend this year is expected to be $2.59T (chips, infra, etc)
There's a very low chance of OpenAI doing an IPO this year: https://polymarket.com/event/openai-ipo-by. So the assumption that they "have to do an IPO" already seems questionable. The rest of the analysis also feels pretty hand-wavy. I'm not saying OpenAI is in a strong financial position, but this article doesn't make a solid enough case for why it's supposedly in such dire straits.
are we genuinely using prediction gambling as sources now?
If you find a more informed and up-to-date source than a market where people wagered close to $1M on I'm all ears.
Any particular preparations to do before this ship hits the sand?
Extract as much liquidity as you can and don't have any commitments to them.
"It means that OpenAI could soon rack up losses that exceed its asset value. In other words, they are on course to slam into bankruptcy this year."
...that's not how bankruptcy happens. What is this guy's background?
Assets on a balance sheet are held at book value. You can absolutely run GAAP losses that exceed net assets without running into bankruptcy, particularly if you're granting (and having employees exercise) options.
The critical measures are cash in and out and debt-like obligations. None of those metrics point to OpenAI going bankrupt this year unless they do something really fucking creative. (Which, to be clear, is Altman's M.O.)
You're right of course, but in this case the statement doesn't look that false. The source article [0] states that half of the assets is cash, so "assets" is defined a bit more freely here.
If we assume a 2026 revenue of 23B (Q1 2026 revenue x4) and costs of 70B (2025 costs * (2026 revenue / 2025 revenue)), even those 25B in cash reserves won't mean much. And this napkin math even ignores their debt obligations. So either they raise money in the range of 50-75B this year (actual money, not datacenter vouchers), or ...
[0] https://www.wheresyoured.at/exclusive-openai-financials/
> costs of 70B (2025 costs (2026 revenue / 2025 revenue)), even those 25B in cash reserves won't mean much*
If all those costs are cash costs, sure. When OpenAI gets non-cash investment, and it "burns" that investment on compute, that should be counted as a GAAP expense. Yet it doesn't touch cash. (It would destroy a unit of compute asset. But again, book versus market value can mess with how that works intuitively.)
Similarly, if you're granting lots and lots of options they're going to generate lots and lots of compensation expenses (and thus losses) as they vest. These aren't cash expenses, however.
Using "on pace to miss its own forecast" is a weird way of phrasing "OpenAI made it up to bullshit the rubes".
Not to mention that being "on pace" is usually a term that means "keeping pace with".
People heavily relying on LLM literally live in a Truman show.
A simulacrum, simultaneously more detailed than the ones built by Meta and Instagram and Tik Tok, but even more shallow, without even the illusion of human connection.
An endless forest of mirrors reflecting no one but the user.
What a shitty title.
“An analyst thinks OpenAI may miss its 2030 ad revenue target by 90%” is what the article says.
[flagged]
[flagged]
the ad business is dead and openai's brand is toxic.
The ad business will never die. It’s capitalisms air and water.
True, but the current state of advertising is, charitably, kinda goofy.
I'm not surprised the iteration we're seeing now is perceived as failing: on the few remaining screens where ads are still present for me, they're between highly irrelevant to flat out repulsive. There are a few brands I will never touch with a long stick, purely as a result of their disturbing, disgusting ads -- again, charitably, a negative-sum game.
As someone who doesnt really see ads either, we are in the minority. The vast majority of people are bombarded with ads from their TVs, streaming platforms, and unadblocked browsers. Try using a relatives iPhone or something its mad.
[dead]
[dead]