The author observes that a call to GPT-5.6 Luna is only 4-5 orders of magnitude more expensive than grep, and then predicts that at current rates of progress, calling an LLM will soon be cheaper than a grep. I think this is a good time to invoke Stein's Law: "If something cannot go on forever, it will stop." These efficiency improvements won't continue forever. It's more likely that the per-call cost of high-quality, compiled software like grep will be a lower-bound that LLMs asymptotically approach, rather than a line that they blow past with perpetual exponential progress. (Barring a true breakthrough in something like quantum computing or room-temperature superconductors.)
You can burn anything* into an ASIC to make it cheaper per-call.
non-backreferencing grep is not very difficult to implement in an ASIC either. But it's probably not worth it because of how relatively rarely you use it and of the data transfer costs.
LLMs are great candidates for ASIC-burning because they're slow compared even to network speeds and run all the time. The issue is that you don't want to burn a specific model or architecture that then becomes obsolete.
So you've got two possible futures, and both guarantee large price drops: (a) LLMs keep getting better and better and better, so ability/$ keeps rising; or (b) LLMs plateau in ability, in which they will start getting ASIC'd.
> LLMs plateau in ability, in which they will start getting ASIC'd.
They don't need to plateu for that to happen. There are companies already building AI on ASIC, and IIRC they were approach 12 months lead time. A 12 months old frontier model (Sonnet 4.5, GPT-5, Kimi K2) for 1% of the price is still a rather good value proposition.
This whole story really reminds me of crypto coins. Like.. going from mining one coin, or lets say token, to millions of fractions like 0.00000000001 bitcoin a week.
depends on what you are grepping ... greapping a large file might be more expensive one day than generating n-th token with LLM that works fully in hardware
you could make hardware implementation of grep and store the file itself next to it in some ROM but that's not a very useful grep ... while hardware LLM is exactly as useful as software LLM only orders of magnitude faster
LLM is spicy memoizing, so it can potentially be faster than a tool call. But people will spend a month tweaking and testing to ensure they have the level of determinism they need, which means it's more expensive, and that they should have used actual memoization in the first place.
It might never beat out grep, but it could beat some more expensive to call tools, similar to how heuristics will often be faster than exact answers. Rust Analyzer can be slow at times, I could see an AI tool taking over a subset of its work.
From a computational standpoint this is obviously nonsense, but from an attentional one I'm not so sure. It may already be more attentionally expensive to use grep in some cases, such the moment you need to remember a non standard arg. And if this applies for performing a simple http operations, then it certainly applies going up the complexity chain.
It's true that LLMs "want" to be be local, but they won't shift broadly to being local until there's a sufficiently large supply of VRAM or (at least) "unified" memory from the manufacturers. (I'm also assuming here that radical regulatory changes like government bans of local models aren't going to happen.) So (AFAICS—I am no expert) the future of LLMs over the next few years comes down primarily to the nitty-gritty of how much memory fab capacity will be added and when, and to a lesser extent of what happens to future demand from LLM SaaS services (& maybe their existing stock of hardware if they get in trouble). (I'm also assuming no roughly-AGI-sized leap forward which makes the frontier models of the near future vastly more valuable than the near-fontier models of today.) For the incumbent manufacturers the high-margin business is selling to LLM SaaS providers who use VRAM efficiently, but the high-volume business is getting chips into millions of laptops which will use VRAM very inefficiently. I assume that they will want to move from high margins to high volumes as they build they physical capacity to ship higher volumes, but they seem to prefer to do it at a stately pace. Hopefully some jostling from Chinese competitors, and maybe a dropoff in demand from data centres, will speed things along.
I found the OP insightful and worth a read. Thank you for sharing it on HN.
The only aspect that is poorly analyzed by the OP is business model viability. All players are investing insane amounts of money in infrastructure with the expectation that their future profits will justify all that investment. The winner or winners in the AGI race, they believe, will find the proverbial "pot of gold at the end of the rainbow."
The OP glosses over questions of business model viability with a brief qualitative discussion and very little hard data. For example, to earn an annual return > 10% on every trillion dollars of capital sunk into infrastructure, the owners of that infrastructure must earn free cash flow (operating profit less investment) in excess of $100 billion per year in perpetuity. Is that feasible? Why? How?
I think the article's analysis is basically right in a vacuum. That is, I think it's clear that inference is a viable business model. But what isn't clear is whether it will be such a profitable business model for any given company that it will justify the investment that company has taken. I kind of think the winners might be a follow-on generation of companies that focus on this commodity inference business model instead of the invent-machine-god-first "business model" and thus are wiser about their level of investment and capital costs.
> I think it's clear that inference is a viable business model.
You may be right. I'm not so sure. Inference looks like a viable business model for those operators that have SOTA infrastructure in place, but the investment required to have it is enormous, and appears to be never-ending, because if an operator stops investing aggressively, its infrastructure quickly becomes non-competitive, and customers will quickly leave for alternatives. SOTA infrastructure is a moving target.
Some frontier labs are reporting positive "adjusted EBITDA" (earnings before interest, taxes, depreciation, and amortization, with extra adjustments to make the figure positive).
Free cash flow (operating profit less investment), actual cash coming in, is deeply in the red.
EBITDA can be a sensible measure of profitability when there isn't much need for additional investment. That doesn't seem to be the case with these operators. They need to invest aggressively to avoid losing customers to competitors. All of these operators have made multi-year commitments to invest more in infrastructure. In addition, they have guaranteed quite a bit of debt to fund it.
Maybe it all will work out fine (and I sure hope it does!), but I didn't see any hard data from the OP, or from you, supporting that view.
EBITDA might make sense for the resellers who package up open weight models and sell inference. It is not appropriate for the labs who have billions in debt for RAM, new data centers, gobbling up competitors, etc.
Those real debt obligations are going to want to be paid back.
Definitely. The question is: Is it enough to recoup the enormous capital costs and justify the level of investment they've received. I think there's a decent chance that it will be. But maybe not. And the longer they keep focusing on training new models more so than on inference, the more uncertain I become that it's all going to work out.
Exactly. "Cost-to-distill" is a critical parameter. Right now usage of frontier models for all tasks is both subsidized and irrationally popular even at the subsidized price. Deepseek would solve most tasks faster and 10x cheaper. I agree with the author that just as Deloitte exists, frontier labs will exist. But not because their products are proprietary technical marvels or gods, but rather because of branding.
DS wouldn't be 10x cheaper than the subsidized subscription plans from openai/anthropic. Although it is of course much cheaper than the enterprier/API pricing- I think if you're on the subscription plans, you can't beat that on performance per price.
It's a deliberate reference/meme that is basically used to acknowledge the precedent of overly exuberant predictions of cost in an emerging technology but argue "however, this time it's true".
Of course, perilous territory for future irony depending on how your prediction plays out.
It's really quite unfortunate that the promise was not delivered, mostly for political reasons. I hope that a new wave of reactors and the dire need for clean energy restarts the nuclear race.
The economics are not there. Solar + batteries are much cheaper per watt today and still improving. Even better, the solar can come online instantly and expand while US nuclear takes twenty years to start generating any energy.
I think Nvidia is under the same pressure as Anthropic/OpenAI. Nvidia will dominate research and probably keep dominating training, but the real volume is in inference. And for inference Nvidia's lead is only a few months, similar to the lead frontier labs have over open source. Nvidia will sell a lot of Rubin CPX's, but their margin on that will be a lot smaller than B200 because there is so much more competition in that space.
There seems to be a mistake in the cost comparison between 2025 and 2026. The 2025 chart axis is the cost to run the entire "intelligence index", and the 2026 version is a weighted average cost per task.
I don't disagree with the thesis here, I just don't think costs are coming down quite that quickly.
A much deeper analysis on the falling price per task was published yesterday by Epoch AI [1]. It's a real statistical analysis and comes to more defensible and grounded conclusions. The headline takeaway is:
The cost of a given level of performance often falls fastest right after that level is first achieved, that is, when it is state of the art (SOTA). We see this pattern on three of our five main benchmarks of AI capability. Averaging across all five, cost falls 66% per quarter (75× per year) for performance that has just debuted as SOTA. Two years later, prices fall half as fast, at 32% per quarter (4.7× per year).
but the analysis itself has more nuance and is a quite interesting read.
for the first chart (sourced from https://epoch.ai/data/machine-learning-hardware?view=graph&y...), what is the audience supposed to think about that trend line? there's a step after you slap a regression on some points where you evaluate whether there's a real trend or noise, right? i don't see that either in the article or the linked source.
GPU case seems very weak. The graph is impossible for me to reason about at least. You could draw basically any trend line through that GPU graph and it would look equally plausible to me. The main takeaway I get is that the NVidia H100 from four whole years ago is barely different in efficiency from the state of the art, which is surprising to me, and seems to indicate the exact opposite of what the article says.
I think this is a case where just drawing a "line goes up" extrapolation is incredibly misleading because there is _tremendous_ economic pressure to get costs down, and costs are very tightly tied to energy use. All of these systems are incredibly inefficient right now and have a lot of room to go down in energy use. I'd guess that the absolute _floor_ is burning model weights directly to silicon and that's like a 90+% reduction in energy use.
Not really related to the central point, by but I couldn't help but get caught up by
> Generally, models intended to be run locally will be much smaller, such as Muse Glimmer or Qwen3 Coder.
That is such an interesting set of models to use as examples here. One being essentially obsolete on release a month ago, and the other being completely ancient in LLM time. I really wonder how they landed on those two.
They're not secretly subsidizing, they're openly subsidizing.
Token pricing was a small minority of customers up until this year, when all the labs started trying to force customers onto token-based billing. Within the last week, Anthropic repriced my team's plan from a temporary "50% extra tokens" to 25%: https://support.claude.com/en/articles/15910845-claude-code-...
The fact that all this is ongoing within such a short timeframe should make you suspicious of any analysis that claims to be observing "statistical trends" like they've discovered a new Moore's Law out of 6 months of pricing data from 2 companies.
The article observes that the cost of frontier intelligence from 2025 has fallen 100x in the last year. It also notes that the energy to run models is also collapsing. Consumer hardware is borked right now because these new algorithms are revolutionizing the utility of a computer. Computing is technology who's cost has been collapsing for 90 years, and its a safe prediction that it will decrease again.
> Yeah okay bud, anyone checked in with the state of consumer hardware recently? Not the author, evidently.
RAM prices will crash when demand drops even a little. They'll probably crash to a lower (inflation adjusted) level than before. This has happened before.
Industrial scaling in general often looks like a sawtooth: price spike, capacity investment, crash, repeat.
Part of what's keeping prices high a little longer is that everyone knows this and is a little reluctant to plow resources into chip fabs for fear of having the bottom fall out before they recoup or sell that to someone else to hold that bag.
Graph the average compute and RAM in a mid-high end laptop at an inflation adjusted price point for the past 40 years. It's very exponential and hasn't slowed down much.
No. Prices will crash when supply side expands to meet the increased demand. Because demand won't go down to pre-bubble times any time soon. Unfortunately the supply side has been very slow in increasing production, partly because most steps of the production chain are all maxed out.
On a long enough scale you are right that prices will likely normalize to a better level, but before 2030? That would mean the factories are built quickly once they begin.
This is the core of my belief that data center construction is a huge bubble.
AI is not a bubble, IMO, though we may see a retrench and some companies with sky-high valuations will crash to more reasonable ones. But data center demand is probably a bubble, and the main driver will be reduction in the actual amount of power and data center space required to serve escalating demand.
I think hardware and model improvements will pace or maybe outrun demand and then when demand starts to saturate will keep going and leave a lot of orphaned data centers.
Jevon's paradox isn't a physical law, it doesn't magically apply to everything. Millions more copies of Atari's ET game didn't cause everyone to pickup a cheap copy, and cause extra demand for a garbage video game. Some times (actually, usually, I'd argue) things are made that will sell for less than the cost of construction because of irrationality, and they don't induce extra demand and they don't change the negative profit margins.
You can't simply wave Jevon's paradox at things. Thousands of miles of canals were dug in the UK that couldn't be sustained and were abandoned. Thousands of miles of railways were laid that could be sustained and were abandoned. And those are potentially durable investments, unlike cheap walls, pillars and roofs laid over a levelled concrete slab full of fast depreciating IT equipment.
I'm so glad the tide here is turning on this talking point, brought on by exactly the same people beating us over the head with it for months while no progress is made towards it materializing.
Many, many people who post here are capable neither of real analysis nor distinguishing real analysis from memes. They aren't hackers, they are adherents of a cult that happens to focus on the same subject matter as hackers.
> Tokens become cheaper than tool calls
The author observes that a call to GPT-5.6 Luna is only 4-5 orders of magnitude more expensive than grep, and then predicts that at current rates of progress, calling an LLM will soon be cheaper than a grep. I think this is a good time to invoke Stein's Law: "If something cannot go on forever, it will stop." These efficiency improvements won't continue forever. It's more likely that the per-call cost of high-quality, compiled software like grep will be a lower-bound that LLMs asymptotically approach, rather than a line that they blow past with perpetual exponential progress. (Barring a true breakthrough in something like quantum computing or room-temperature superconductors.)
Yeah I bumped on that too. If it's possible to make llms cheaper than current grep, then it is also almost certainly possible to make grep cheaper.
You can burn anything* into an ASIC to make it cheaper per-call.
non-backreferencing grep is not very difficult to implement in an ASIC either. But it's probably not worth it because of how relatively rarely you use it and of the data transfer costs.
LLMs are great candidates for ASIC-burning because they're slow compared even to network speeds and run all the time. The issue is that you don't want to burn a specific model or architecture that then becomes obsolete.
So you've got two possible futures, and both guarantee large price drops: (a) LLMs keep getting better and better and better, so ability/$ keeps rising; or (b) LLMs plateau in ability, in which they will start getting ASIC'd.
> LLMs plateau in ability, in which they will start getting ASIC'd.
They don't need to plateu for that to happen. There are companies already building AI on ASIC, and IIRC they were approach 12 months lead time. A 12 months old frontier model (Sonnet 4.5, GPT-5, Kimi K2) for 1% of the price is still a rather good value proposition.
This whole story really reminds me of crypto coins. Like.. going from mining one coin, or lets say token, to millions of fractions like 0.00000000001 bitcoin a week.
depends on what you are grepping ... greapping a large file might be more expensive one day than generating n-th token with LLM that works fully in hardware
you could make hardware implementation of grep and store the file itself next to it in some ROM but that's not a very useful grep ... while hardware LLM is exactly as useful as software LLM only orders of magnitude faster
LLM is spicy memoizing, so it can potentially be faster than a tool call. But people will spend a month tweaking and testing to ensure they have the level of determinism they need, which means it's more expensive, and that they should have used actual memoization in the first place.
Right. And some hardware improvements will speed up both grep and Luna, which won't close the gap.
not necessarily, one may be easier to parallelize while the other suffers some serial computation bottleneck.
Grep is trivially parallelizable if you care to do it though.
https://iepathos.github.io/ripgrep/performance/#work-stealin...
It might never beat out grep, but it could beat some more expensive to call tools, similar to how heuristics will often be faster than exact answers. Rust Analyzer can be slow at times, I could see an AI tool taking over a subset of its work.
> calling an LLM will soon be cheaper than a grep
From a computational standpoint this is obviously nonsense, but from an attentional one I'm not so sure. It may already be more attentionally expensive to use grep in some cases, such the moment you need to remember a non standard arg. And if this applies for performing a simple http operations, then it certainly applies going up the complexity chain.
It's true that LLMs "want" to be be local, but they won't shift broadly to being local until there's a sufficiently large supply of VRAM or (at least) "unified" memory from the manufacturers. (I'm also assuming here that radical regulatory changes like government bans of local models aren't going to happen.) So (AFAICS—I am no expert) the future of LLMs over the next few years comes down primarily to the nitty-gritty of how much memory fab capacity will be added and when, and to a lesser extent of what happens to future demand from LLM SaaS services (& maybe their existing stock of hardware if they get in trouble). (I'm also assuming no roughly-AGI-sized leap forward which makes the frontier models of the near future vastly more valuable than the near-fontier models of today.) For the incumbent manufacturers the high-margin business is selling to LLM SaaS providers who use VRAM efficiently, but the high-volume business is getting chips into millions of laptops which will use VRAM very inefficiently. I assume that they will want to move from high margins to high volumes as they build they physical capacity to ship higher volumes, but they seem to prefer to do it at a stately pace. Hopefully some jostling from Chinese competitors, and maybe a dropoff in demand from data centres, will speed things along.
I found the OP insightful and worth a read. Thank you for sharing it on HN.
The only aspect that is poorly analyzed by the OP is business model viability. All players are investing insane amounts of money in infrastructure with the expectation that their future profits will justify all that investment. The winner or winners in the AGI race, they believe, will find the proverbial "pot of gold at the end of the rainbow."
The OP glosses over questions of business model viability with a brief qualitative discussion and very little hard data. For example, to earn an annual return > 10% on every trillion dollars of capital sunk into infrastructure, the owners of that infrastructure must earn free cash flow (operating profit less investment) in excess of $100 billion per year in perpetuity. Is that feasible? Why? How?
The OP does not really consider such questions.
I think the article's analysis is basically right in a vacuum. That is, I think it's clear that inference is a viable business model. But what isn't clear is whether it will be such a profitable business model for any given company that it will justify the investment that company has taken. I kind of think the winners might be a follow-on generation of companies that focus on this commodity inference business model instead of the invent-machine-god-first "business model" and thus are wiser about their level of investment and capital costs.
> I think it's clear that inference is a viable business model.
You may be right. I'm not so sure. Inference looks like a viable business model for those operators that have SOTA infrastructure in place, but the investment required to have it is enormous, and appears to be never-ending, because if an operator stops investing aggressively, its infrastructure quickly becomes non-competitive, and customers will quickly leave for alternatives. SOTA infrastructure is a moving target.
A phrase comes to mind: "Your margin is my opportunity."
They are already turning profits and inference has shown to be a cash cow. And they've already secured compute for the next several years.
Some frontier labs are reporting positive "adjusted EBITDA" (earnings before interest, taxes, depreciation, and amortization, with extra adjustments to make the figure positive).
Free cash flow (operating profit less investment), actual cash coming in, is deeply in the red.
EBITDA can be a sensible measure of profitability when there isn't much need for additional investment. That doesn't seem to be the case with these operators. They need to invest aggressively to avoid losing customers to competitors. All of these operators have made multi-year commitments to invest more in infrastructure. In addition, they have guaranteed quite a bit of debt to fund it.
Maybe it all will work out fine (and I sure hope it does!), but I didn't see any hard data from the OP, or from you, supporting that view.
EBITDA might make sense for the resellers who package up open weight models and sell inference. It is not appropriate for the labs who have billions in debt for RAM, new data centers, gobbling up competitors, etc.
Those real debt obligations are going to want to be paid back.
Definitely. The question is: Is it enough to recoup the enormous capital costs and justify the level of investment they've received. I think there's a decent chance that it will be. But maybe not. And the longer they keep focusing on training new models more so than on inference, the more uncertain I become that it's all going to work out.
Who is the "they" that are turning profits?
Exactly. "Cost-to-distill" is a critical parameter. Right now usage of frontier models for all tasks is both subsidized and irrationally popular even at the subsidized price. Deepseek would solve most tasks faster and 10x cheaper. I agree with the author that just as Deloitte exists, frontier labs will exist. But not because their products are proprietary technical marvels or gods, but rather because of branding.
DS wouldn't be 10x cheaper than the subsidized subscription plans from openai/anthropic. Although it is of course much cheaper than the enterprier/API pricing- I think if you're on the subscription plans, you can't beat that on performance per price.
Too Cheap to Meter reminds me of the promise of Nuclear Power in 1954
"It is not too much to expect that our children will enjoy in their homes electrical energy too cheap to meter,..." Lewis Strauss
https://en.wikipedia.org/wiki/Too_cheap_to_meter#Origins
Oddly enough my power bill was metered and big.
It's a deliberate reference/meme that is basically used to acknowledge the precedent of overly exuberant predictions of cost in an emerging technology but argue "however, this time it's true".
Of course, perilous territory for future irony depending on how your prediction plays out.
It's really quite unfortunate that the promise was not delivered, mostly for political reasons. I hope that a new wave of reactors and the dire need for clean energy restarts the nuclear race.
The economics are not there. Solar + batteries are much cheaper per watt today and still improving. Even better, the solar can come online instantly and expand while US nuclear takes twenty years to start generating any energy.
> NVIDIA will still boom
I think Nvidia is under the same pressure as Anthropic/OpenAI. Nvidia will dominate research and probably keep dominating training, but the real volume is in inference. And for inference Nvidia's lead is only a few months, similar to the lead frontier labs have over open source. Nvidia will sell a lot of Rubin CPX's, but their margin on that will be a lot smaller than B200 because there is so much more competition in that space.
There seems to be a mistake in the cost comparison between 2025 and 2026. The 2025 chart axis is the cost to run the entire "intelligence index", and the 2026 version is a weighted average cost per task.
I don't disagree with the thesis here, I just don't think costs are coming down quite that quickly.
A much deeper analysis on the falling price per task was published yesterday by Epoch AI [1]. It's a real statistical analysis and comes to more defensible and grounded conclusions. The headline takeaway is:
The cost of a given level of performance often falls fastest right after that level is first achieved, that is, when it is state of the art (SOTA). We see this pattern on three of our five main benchmarks of AI capability. Averaging across all five, cost falls 66% per quarter (75× per year) for performance that has just debuted as SOTA. Two years later, prices fall half as fast, at 32% per quarter (4.7× per year).
but the analysis itself has more nuance and is a quite interesting read.
[1] https://epoch.ai/publications/the-plunging-price-of-thought
for the first chart (sourced from https://epoch.ai/data/machine-learning-hardware?view=graph&y...), what is the audience supposed to think about that trend line? there's a step after you slap a regression on some points where you evaluate whether there's a real trend or noise, right? i don't see that either in the article or the linked source.
GPU case seems very weak. The graph is impossible for me to reason about at least. You could draw basically any trend line through that GPU graph and it would look equally plausible to me. The main takeaway I get is that the NVidia H100 from four whole years ago is barely different in efficiency from the state of the art, which is surprising to me, and seems to indicate the exact opposite of what the article says.
It is hard to see how the environmental side effects of this aren't going to be somewhere between bad and disastrous.
I think this is a case where just drawing a "line goes up" extrapolation is incredibly misleading because there is _tremendous_ economic pressure to get costs down, and costs are very tightly tied to energy use. All of these systems are incredibly inefficient right now and have a lot of room to go down in energy use. I'd guess that the absolute _floor_ is burning model weights directly to silicon and that's like a 90+% reduction in energy use.
No it'll be fine as long as you do your part and not drive a car, or have AC, or eat meat, or have children, or live in detached housing, or...
Not really related to the central point, by but I couldn't help but get caught up by
> Generally, models intended to be run locally will be much smaller, such as Muse Glimmer or Qwen3 Coder.
That is such an interesting set of models to use as examples here. One being essentially obsolete on release a month ago, and the other being completely ancient in LLM time. I really wonder how they landed on those two.
We have people suggesting that ai is so costly to run that all labs are secretly subsidising tokens and we can expect a reprice soon.
Then we have these articles that say tokens will get so cheap that labs won’t know how to make profit.
Who is correct?
They're not secretly subsidizing, they're openly subsidizing.
Token pricing was a small minority of customers up until this year, when all the labs started trying to force customers onto token-based billing. Within the last week, Anthropic repriced my team's plan from a temporary "50% extra tokens" to 25%: https://support.claude.com/en/articles/15910845-claude-code-...
The fact that all this is ongoing within such a short timeframe should make you suspicious of any analysis that claims to be observing "statistical trends" like they've discovered a new Moore's Law out of 6 months of pricing data from 2 companies.
So this week's coordinated shock and awe campaign is about "open models" and "prices are falling rapidly". Also on the front page:
https://news.ycombinator.com/item?id=49815526
>We are likely to see LLMs running locally at current frontier-quality on commodity hardware in the next 3-6 years
Yeah okay bud, anyone checked in with the state of consumer hardware recently? Not the author, evidently.
>oh in 3-6 years this will all be over
Yeah I'm sure Samsung, Nvidia and sk hynix will all be very calm with lower volumes and lower margins.
Seems like 3-6 years is a pretty conservative timeline to me. Memory and interview chip production could ramp up massively in that amount of time.
The article observes that the cost of frontier intelligence from 2025 has fallen 100x in the last year. It also notes that the energy to run models is also collapsing. Consumer hardware is borked right now because these new algorithms are revolutionizing the utility of a computer. Computing is technology who's cost has been collapsing for 90 years, and its a safe prediction that it will decrease again.
> Yeah okay bud, anyone checked in with the state of consumer hardware recently? Not the author, evidently.
RAM prices will crash when demand drops even a little. They'll probably crash to a lower (inflation adjusted) level than before. This has happened before.
Industrial scaling in general often looks like a sawtooth: price spike, capacity investment, crash, repeat.
Part of what's keeping prices high a little longer is that everyone knows this and is a little reluctant to plow resources into chip fabs for fear of having the bottom fall out before they recoup or sell that to someone else to hold that bag.
Graph the average compute and RAM in a mid-high end laptop at an inflation adjusted price point for the past 40 years. It's very exponential and hasn't slowed down much.
No. Prices will crash when supply side expands to meet the increased demand. Because demand won't go down to pre-bubble times any time soon. Unfortunately the supply side has been very slow in increasing production, partly because most steps of the production chain are all maxed out.
On a long enough scale you are right that prices will likely normalize to a better level, but before 2030? That would mean the factories are built quickly once they begin.
> everyone knows this and is a little reluctant to plow resources into chip fabs
except cxmt who is plowing resources in like crazy
That has more to do with geopolitics than it does the current price of memory.
This is the core of my belief that data center construction is a huge bubble.
AI is not a bubble, IMO, though we may see a retrench and some companies with sky-high valuations will crash to more reasonable ones. But data center demand is probably a bubble, and the main driver will be reduction in the actual amount of power and data center space required to serve escalating demand.
I think hardware and model improvements will pace or maybe outrun demand and then when demand starts to saturate will keep going and leave a lot of orphaned data centers.
> AI is not a bubble
When people say "AI is a bubble", they mean economically as a whole, which includes data centers.
Perhaps we need better terminology for "product useful; numbers nonsensical"
Jevon's paradox says that if data centers can serve a lot more tokens per dollar or watt there will be increased demand for data centers.
Jevon's paradox isn't a physical law, it doesn't magically apply to everything. Millions more copies of Atari's ET game didn't cause everyone to pickup a cheap copy, and cause extra demand for a garbage video game. Some times (actually, usually, I'd argue) things are made that will sell for less than the cost of construction because of irrationality, and they don't induce extra demand and they don't change the negative profit margins.
You can't simply wave Jevon's paradox at things. Thousands of miles of canals were dug in the UK that couldn't be sustained and were abandoned. Thousands of miles of railways were laid that could be sustained and were abandoned. And those are potentially durable investments, unlike cheap walls, pillars and roofs laid over a levelled concrete slab full of fast depreciating IT equipment.
It's true that Jevon's paradox doesn't always apply, although this does seem like a classic case.
But yes, if sold for a negative margin Jevon eventually stops because the decreasing supply will drive up prices.
> things are made that will sell for less than the cost of construction
Price is set at the marginal cost. Capital costs aren't in marginal costs.
You'll need a better counter-example than UK railways which suffered from Parliament price-fixing.
> You can't simply wave Jevon's paradox at things
I'm so glad the tide here is turning on this talking point, brought on by exactly the same people beating us over the head with it for months while no progress is made towards it materializing.
Many, many people who post here are capable neither of real analysis nor distinguishing real analysis from memes. They aren't hackers, they are adherents of a cult that happens to focus on the same subject matter as hackers.
Jevon's Paradox, RSI, revealed preference
I agree, but I would say what the parent is saying is more akin to Say's law: https://en.wikipedia.org/wiki/Supply_creates_its_own_demand