Again, Zitron provides the numbers for his theories and they’re very compelling.
I’m still not seeing any equally compelling arguments as to why this is not the case. Only accusations of doomerism and links to him calling the bubble collapse early.
> I’m still not seeing any equally compelling arguments as to why this is not the case.
Zitron relies too heavily on how big the numbers are and not how workable the numbers are. Further, he seems to think that it will all just implode, which is pretty unlikely.
AI companies are making money. 1T in purchase negotiations is something that can be renegotiated if the numbers don't improve. And there's actually a pretty good chance that these AI companies sell the US federal government on AI being a strategic advantage which can ultimately gets a nice federal funding source.
Even in the worst case of what ed predicts, the more likely outcome is that the AI companies slow rollouts and purchases. The general market takes a hit, but it's ultimately not the end of the world.
But further, even with AI reducing their consumption, that doesn't mean chip manufacturers are hosed, we've already built up huge demand for things like RAM which are supremely supply constrained. That' has slowed the sale of consumer and enterprise electronics. Easing back on the AI market means those markets will likely pick up the slack again. Especially because I suspect businesses will be seriously thinking about things like "Why don't we deploy deepseek locally to save on compute cost?".
I suspect that prices for AI will ultimately increase before any of this happens and with those price increases that's where I can see there being more a demand to break ties with the bigger AI companies.
What Ed misses is that big business has much MUCH more flexibility when it comes to financing than even a midsize corperation. They have direct lines to bank presidents.
I agree the issue is not the scale but you're missing the reason the scale has allowed to inflate to these proportions is through purchase agreements that are enabled by other purchase agreements. Yes banks, other funders, and even the cash flow of the largest companies can provide lifelines, but why would they if the pullback begins? The problem with circular purchase agreements is the mechanism that allowed them to grow rapidly would also force them to shrink rapidly and most funders aren't going to want to catch that falling knife unless profit is near guaranteed, which it isn't. The numbers are still deeply in the world of speculative. Doesn't matter if you have a line to a bank president, they don't want to bail you out for free.
The point is that the numbers as they exist now require the industry to shoot the moon. It's not impossible, maybe the technology really is that revolutionary. But the margin for error is microscopic.
Let's say a kid's lemonade stand increases its revenue from $1/month to $100/month in three months. Clearly that 100x revenue increase per quarter is sustainable, so it predicts a $100M revenue a year from now, raises money at a $1B valuation, and signs a contract to buy $10M in industrial lemon squeezing machines. The squeezer manufacturer then predicts that it'll see a $500M revenue from all the other lemonade companies, raises money, and spends $1B on expanding its factories. Meanwhile, the kid is actually losing money because they are selling gold-leaf lemonade for $1 / glass while it costs $100 / glass to produce.
But the lemonade stand is making money! Worst-case scenario they'll just renegotiate the squeezer contract and slow down their growth, right? If the lemonade business goes bust they'll just sell hundreds of millions of dollars of squeezers to lime juice stands, right? The government will declare lemonade a "resource critical for national defense" and bail out the industry, right?
Yeah, sure, AI isn't 100% bullshit. There is indeed some money there, it won't be a complete collapse. But the numbers we are seeing are absolutely insane. We're already at "AI is bigger than the entire internet" levels of investments! Either we are collectively burning hundreds of billions of dollars on pure hype, or somehow every single company on the planet is hiding a secret 10x AI-boosted productivity increase.
Be careful that you are not basing your argument on whether you personally like lemonade.
If a kid's lemonade stand increased its revenue from $1B/year to $100B/year in two years, we would be paying serious attention to their product. People seem to love that lemonade! At this level of increased demand, maybe it's ok to let this kid borrow $1T over 5 years, because even if his growth slows down to 2x/year instead of 10x/year, he will still be covering that investment. That's hopeful and risky, but not at all insane in my view.
And what's going to happen to those hundreds of billions of investments in equipment that won't be needed until well after it has become obsolete and uneconomical to operate?
What happens to the stock price of a company that missed its revenue targets by a factor of 5?
Making extreme growth predictions and not realizing them is fine for small startups. Not so much when it comes to trillion-dollar companies - especially when they have taken on massive loans to pay for it.
A 2x-instead-of-10x result? That's an economy-killing crash.
> What happens to the stock price of a company that missed its revenue targets by a factor of 5?
It depends. Tesla is a good example of how those things are not as clear cut as you might think.
The world is starving for more AI compute. Nvidia's GPUs are selling for 2x or 3x their former price, RAM prices increased even more, all because demand is huge right now. If anything buying those chips at bulk discount and then selling them right back to consumers would already be profitable.
Nobody can rebut 10k words of rambling across half a dozen of separate arguments in a forum post. It'd be too long, and there is no readership for a point-by-point rebuttal for the dozens logic errors, misreprentations and various sleights of hand that the arguments (such as they are) are built on. And if you try to just rebut one thing, well, that wasn't actually the core argument but just incidental.
I've seen people talk about how compelling they find these articles, but literally never have they been able to point at a good argument. Like, in a paragraph, what's the most compelling and impactful argument in this article? What are a couple of numbers that make it so?
By 2029, according to their own projections:
OpenAI ≈ $184B/year
Anthropic ≈ $174B/year
Combined ≈ $358B/year
And remember, this is also in competition with open models that are increasingly encroaching on SOTA capabilities while costing less and allowing you to do more.
The immediate fallacy here is implying they need to make a profit to pay for compute. Profit is what's left over after you pay for expenses, not what you pay the expenses with.
The sleight of hand (yours) is that those numbers don't seem to be from the article. Like, none of those numbers appears in the article except the years, and the years are used in different contexts.
Likewise what you say is the main point ("the labs can't grow fast enough to hit these targets") is not something the article makes an argument about!
In fact, the only even related thing in the article is an this assertion:
> To be clear, “growing super fast” is no longer sufficient for OpenAI and Anthropic.
Which is obviously nonsense, but also at cross-purposes with what you say.
So neither what you say is the main point nor the numbers is in the article at all. Seems pretty sad as demonstration of how compelling the article was.
Ah, forgive me, I’d thought with the way you were attacking his work that you had more of an idea of what his consistent line of reasoning was.
The numbers I’ve provided are the basis for just about every article he has written regarding the AI industry with his main point being that the spend is unjustified and unsustainable and more importantly unwarranted considering the usefulness and returns (which are not small but in no means justify the enormous estimates of profitability being assumed by these companies).
So, no sleight of hand, just an assumption that your acidic regard for his work was rooted in having read it.
So, do you care to provide a rebuttal to the article or his main point about profitability? I don’t imagine calling his theory “obviously nonsense” was as far as your reasoning went. Care to expand on it?
> So, do you care to provide a rebuttal to the article
No, I already explained why rebutting the whole article is impractical. I might have been ok doing so for a single argument just as a demonstration, but your unwillingness to point out the most compelling and impactful argument of this article makes it pretty clear there was no point to it. It was 10k words of padding.
> or his main point about profitability
A message ago the main point was about revenue growth. Now it has shifted to profitability. So, no, I'm not going to waste time rebutting an ill-defined, second-hand, moving target argument either.
> I don’t imagine calling his theory “obviously nonsense” was as far as your reasoning went. Care to expand on it?
It's nonsense by definition? The claim is that companies have too low a revenue to pay for their compute spending commitments, and growing super fast can't fix that problem. Super fast revenue growth is exactly what solves the problem of revenue being too low! It's a very strange thing to assert, but it's what he did assert, and it's the only bit of this article that's even close to what you claim is his main point.
("Oh, but super fast growth can't happen because of blah blah blah as explained in some other article", you'll say. Maybe so. But blah blah blah is not what Zitron wrote in this article, in this one he wrote that revenue growth won't solve the problem of low revenue. Rebutting a 10k word ramble is merely impractical. Rebutting an amorphous blob of 1M contradictory words is impossible, there's always going to be some explanation about how what he really meant is somewhere else.)
Haha…So, to be clear, I was giving you the option to respond to his main, consistent argument as a courtesy because you didn’t want to respond point by point.
I see though that you don’t seem to want to do that either. It’s more that you just really don’t want him to be right is what I’m gathering, which is fine honestly.
I’m confused by your distaste for what he’s saying though. The failure of OAI or Anthropic doesn’t mean AI is bad or goes away, just that cheaper/local models turned out to be the right price point. Do you have some specific desire to see these two labs succeed?
You are cleanly demonstrating why people don't respond to these arguments: it's a waste of time for everyone involved because the goalposts always move.
Nah, these goalposts have stayed exactly where they began. I just don’t think anyone here actually has an argument against him.
I’ll be glad to hear any argument at this point but if we don’t have one yet (other than “he’s wrong because I have a side business selling AI products/services”) I don’t think one is coming.
I get the feeling the author had a very simple and clear idea and then spent a lot of time fumbling about trying to present it clearly.
At one point I'm pretty sure there are six paragraphs shortly after each other that are all trying to restate the paragraphs before it. Since nothing seemed to become much clearer I kind of gave up at that point.
This seems to be the standard response to Ed’s work. He presents facts and figures that backup his theories in a painstaking manner and then detractors reply with something like this.
Following HNs guidelines, my most charitable interpretation of your response is that you don’t understand the arguments or numbers behind them and you aren’t just dismissing the arguments because you don’t like them. For that I’m saying you need to feed the article to the LLM of your choice and work back and forth until you get to an understanding of what’s being argued.
> Following HNs guidelines, my most charitable interpretation of your response is that you don’t understand the arguments or numbers behind them and you aren’t just dismissing the arguments because you don’t like them. For that I’m saying you need to feed the article to the LLM of your choice and work back and forth until you get to an understanding of what’s being argued.
You can call it painstaking but I mostly found it painful to read.
I mean you don't need that much text or numbers to argue most of the AI hardware spend seems driven by two companies, who likely need to spend more next year to match projected growth.
But then there's a lot of repetition and calling those two companies unprofitable and unsustainable, which I really hoped he would expand upon but he didn't get that far before I gave up.
That’s Zitron for ya. To be fair to him, he’s kind of forced into this position: he’s now famous for championing a cause (AI is dumb and will never work and the scientists are lying) that is becoming increasingly untenable by the day.
> AI is dumb and will never work and the scientists are lying
I don't see him making that claim, though.
Screwdrivers aren't useless. Screwdrivers have gotten a lot better over the decades. Everyone uses screwdrivers these days. Investing $500 billion into screwdrivers and expecting the screwdriver industry to grow to a double-digit percentage of GDP is still a really stupid idea. Let's face it: Universal Screwdriver[0] isn't going to happen.
Yeah, again this is the typical response from those who would rather stick their heads in the sand than face the reality that the spending is untenable.
They never refute his claims with anything substantial. They just call you “anti ai” or whatever (despite the argument having nothing to do with the technology, just a subset of companies).
Microsoft’s total depreciation and amortization in Q2 2027 was $11B - not clear how much of this is AI related. Apparently they had $34B in AI ARR as of May.
So let’s say their AI capex amortization and revenue are about equal. Not amazing, obviously they’re relying on continued growth, but doesn’t seem like the end of the world?
Compare that to Ed’s framing - Microsoft has $34B in revenue but spent $116B in capex last year to “make it”. They’re doomed!
But that capex spend is to make future revenue. Clearly he assumes demand won’t increase in the future, and that future projected revenue is “fake”. And sure, it definitely might not increase enough to make profitability.
But his whole analysis hinges on that one assumption. The entire article, all the numbers he gish gallops at you, could basically be replaced with “I don’t think AI demand and revenue will increase much beyond today.” Yeah, we know.
We see that companies are already balking at the cost and are increasingly looking at what the actual return of their current spend is, let alone when these prices have been increasing.
Where is the increase in demand going to be coming from?
Especially the increase needed to make this make sense?
My personal hunch is that the “diffusion curve” for AI is slower than most people in this space think. Most businesspeople I talk to have only tried a basic Copilot chat and/or free ChatGPT. Many still haven’t used anything “AI” at all. As more use cases become practicable and cost-effective, more software will include AI, and more people will use AI with or without knowing.
Also: note that the Wall Street analyst estimates Ed cites (and then declares impossible targets) are predictions by serious people with a lot of money at stake. Of course they could be wrong, but they’re not made up.
A lot of that demand could still flow through to AWS, Azure, etc. Maybe not at some startups, but I'd expect bigger companies would still use the hyperscalers even if they move off OAI/Anthropic.
The real problem for the hyperscalers would be demand stalling out entirely (or maybe small local models getting good enough that people don't use the cloud).
The infra expansion of the hyperscalers is predicated on, and financially justified by, the presumed future growth of OpenAI and Anthropic, and their projected revenue. If the latter is not going to happen, I expect OpenAI and Anthropic to go bankrupt, which will trigger a restructuring of the hyperscalers, which should flood the market with either training HW or training capacity at very low cost, and it's not necessarily the hyperscalers that will benefit the most: I believe some of the GPU clouds are using HW leased or bought form NVidia, and running inside second-tier datacenters, not the hyperscalers.
> Also: note that the Wall Street analyst estimates Ed cites (and then declares impossible targets) are predictions by serious people with a lot of money at stake.
I think Ed is wrong. But I have to push back. Wall street analysts are more likely to misrepresent precisely because they have money at stake. Much like ed has a pretty vested interest in saying the sky is falling (that's his brand at this point) the analysts have vested interests in saying everything is fine and keep investing.
We can see similar behaviors with analysts like zero hedge, which every week write a new "the bubble is about to pop" article.
A lot of this, IMO, is similar to a fact about the weather I'm probably misremember from stats. If you always predict "it will be sunny tomorrow" almost anywhere in the world you'll be right something like 80 to 90% of the time (citation needed).
Analysts who always say "things are great and stocks will go up" will be right most of the time. The tricky thing has always been predicting when and if a pop will happen.
Yes, fair point, analysts tend not to be too critical. But they are still developing more detailed models to understand and project things like revenue and profit, so there's some rigor - at least more than company leadership just giving takes or stating platitudes.
If you believe in the notion that the technology will improve to the point of being able to replace a human employee, the increase in demand is going to come from businesses that choose AI employees over human ones. Even if the AI employee costs more than human one, it's like a car vs a horse. The AI employee doesn't get sick, doesn't get into trouble with HR for sexual harassment, never comes into work hung over. You can hire 100 AI employees for a week and them fire them the next and not feel bad about it.
Whether this comes to pass is anyone's guess, but that's the theory.
I would say Ed makes three main claims in this post:
1) Companies are spending a ton on capex for future AI compute
2) Current levels of AI revenue are not enough to recoup that capex spend
3) Revenues won’t increase enough in the future to recoup that capex spend
Almost anyone, bubbler or not, would agree with points 1 and 2. But Ed cites dozens of numbers from different sources to repeat and reinforce them. It feels to me like an effort to overwhelm the reader with data to support his overall argument. That’s what I would call a gish gallop.
The third point is a prediction. He cites a lot of facts and numbers here too, but ultimately whether you believe his prediction is going to depend on your assumptions.
The thing is, I really would love to see a detailed analysis of capex spend and amortization. Capex spent on the future is a big unknown. But the big labs have claimed they are profitable on inference. How much capex was invested to create the capacity to serve current models? How much revenue is coming from serving those models? What does the full profitability picture look like? What does that imply for future demand needs?
> But the big labs have claimed they are profitable on inference.
As the accountants say, profit is an opinion, but cash flow is a fact.
Like, if these companies are profitable on inference (depends on how you count training expenses I suspect), then they shouldn't need to raise more money.
For reference, Anthropic appear to have raised about $130bn, which is a lot. Assuming that OpenAI have raised about the same.
Even to get a 2x return on this investment they'd need to start making about $50bn per annum (profit, not revenue) for 10 years. That's Google level net income, on a very very different business (google's business is much more capital efficient).
I personally find that very unlikely, but we'll see I guess.
> a detailed analysis of capex spend and amortization
This is kinda irrelevant unless they are making money (which the hyperscalers are, and the pure play model companies are not).
Yes, discussion like this is why I'd love to see a deeper analysis of the full value chain. Right now it's mostly speculation (beyond 'gee, everyone sure is spending a lot of money').
Obviously more detailed data isn't easily available to report on, but I'd like to know:
- What are the labs spending on compute specifically to serve models? Are they really profitable "on inference" of existing models? If so, how long is the payback period to recoup their training costs for those existing models only? If not, how much would prices need to rise to be profitable?
- How much capex have the hyperscalers invested in just the compute being used to serve those existing models? Are they making money "on inference" when accounting for just that amortized capex? If so, what are the margins like?
- What share of hyperscalers' AI revenue is from training vs inference? (Presumably training is more dependent on VC investment and inference is more self-sustaining.)
On your last paragraph the simplest answer as to why we haven’t seen that is because they don’t want to show us because it wouldn’t paint a great picture for them.
Regarding Gish galloping, I don’t think you can Gish Gallup in writing. The point as you said is to rapidly overwhelm an opponent. That’s not possible in writing as the points can be argued one by one at the responders leisure.
Thank you, I will bear that in mind. I am mostly familiar with the term from online forums, where I’ve seen it used to refer to other forum posts, blog posts, etc.
> The other problem is the monstrous and abusive marketing campaign from the AI industry itself, and those who use AI on a regular basis. If you are against the consensus that AI will grow ever-larger every single quarter forever, you will be harassed and dogpiled across multiple social media platforms by everyone from AI influencers to actual journalists. The fact that it’s more professionally dangerous to critique the powerful than it is to align with them is disgusting, but I should be clear that these tactics only reinforce that I’m on the right track.
> As Nik Suresh noted in his recent piece, refusing to say that AI is giving you massive productivity benefits will lead to actual professional consequences, because so much is riding on the overall grift about what AI can do (which is much, much less than the boosters will promise). This runs antithetical to productivity or good sense, and everybody involved in it should be both eternally shamed and shunned from any sensible business.
> And while the AI industry and its fandom will claim that people like me are “skeptics” and “haters,” the outright hatred and vitriol that they spew for not falling in line behind a nakedly false narrative built on outright disinformation is disgraceful.
The most charitable way I can read those paragraphs is that he's vaguely alluding to a nebulous conspiracy by a cabal of elites, which evokes a specific ideology that cannot be discussed constructively.
Ed has repeatedly done the "why are people using LLMs?" bit but ignores what well-respected software engineers have actually said about them because he thinks they're AI boosters and they are just lying. At the least, it's an obvious blind spot to explain how the AI industry is what it is despite the numbers he reports arguing otherwise, and many of his industry trend predictions have been incorrect as a direct result.
You are reading way too much into it. The truth is far simpler: when someone's salary depends on AI being a success, you'll have a really hard time getting them to openly say anything negative about AI.
A huge portion of those "well-respected software engineers" work at companies which have pivoted to make AI an essential part of their strategy - because every company has. We're seeing companies where leadership explicitly says that anyone not enthusiastically making AI a core part of their workflow will get fired. It is literally forbidden to doubt AI. Want a promotion? Better start tokenmaxxing - your manager will be looking at your usage stats!
And it's going to be the same at the higher levels. A "CTO of $company says $100M AI investment was a massive waste" headline is obviously unacceptable. You've already burned the money, so why ruin your stock price by being "the only company too incompetent to properly use AI" rather than slowly and silently scaling it down while pretending everything is fine and parroting all the buzzwords your peers are using on LinkedIn?
The actual results are irrelevant. It does not matter if your AI results are mediocre, or "good in some edge cases", or "can speed up some easier work". The market has decided that AI is huge and a massive success, so anyone not adopting it and having extreme successes with it is a failure and will be punished.
There's no need to invent a conspiracy here. Simple self-preservation will do.
Ah, yes. Because someone owns NVDA in their retirement account, they're going to intellectually bankrupt themselves and totally lie about being able to use an LLM to generate code feels faster than writing it manually. No conspiracy at all!
> The most charitable way I can read those paragraphs is that he's vaguely alluding to a nebulous conspiracy by a cabal of elites, which evokes a specific ideology that cannot be discussed constructively.
If the markets/finance people reward you for saying you are all-in on AI, people with much of their compensation dependent on the markets (i.e. executives) will claim that AI is driving amazing results.
Like, I find AI useful in lots of places and use these tools daily, but I don't see how the pure play model companies can deliver a good return for their investors.
Unlike ed though, I do think that we'll find good uses for all the compute and (hopefully) the energy infrastructure being built out for this compute.
I’m still not seeing any equally compelling arguments as to why this is not the case. Only accusations of doomerism and links to him calling the bubble collapse early.