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. Only accusations of doomerism and links to him calling the bubble collapse early.
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.
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.
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.
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.
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.
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?
How are OAI and Anthropic reasonably and reliably going to make the profits they need to to hit these targets?
By Q1 2028: OpenAI ≳ $10B/month Anthropic ≳ $10B/month Combined ≳ $20B/month / $240B ARR
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 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.
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?
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.)
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?
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.