Some data would reinforce your case. Do you have it?
Here is my data point: "You Have No Idea How Screwed OpenAI Actually Is" - https://wlockett.medium.com/you-have-no-idea-how-screwed-ope...
Palm is closer but it's a different world. It's established that Internet advertising companies are worth trillions. It's only in retrospect that what Palm could have been is obvious.
Barring something very unexpected OpenAI is coming out on top. They're prepaying for a good 5-10 years of compute. That means their inference and training for that time are "free" because they've been paid for. They're going to be able to bury their competition in money or buy them out.
OpenAI is also first, but it is absolutely not a given that they are the Apple in this situation. Microsoft too had money to bury the competition, they even staged a fake funeral when they shipped windows phone 7.
> Barring something very unexpected
Like the release of an iPhone?
Blackberry was a big deal for a while, too
Then Apple came out with the iPhone but it was not clear they were going to be the market leader. They were certainly not the first mover.
With OpenAI, I see even less of a moat. Someone else comes along, makes a better foundational model, and they've got a gigantic advantage. What does OpenAI have?
This is where the money is. Anthropic just released claude for excel. If it replaces half of the spreadsheet pushers in the country theyre looking at massive revenue. They just started with coding because theres so much training data and the employees know a lot about coding
The reason I wonder about that is because that also seems to be the dynamic with all these deals and valuations. Surely if OpenAI would pay $30 billion on data centers, they could pay $40 billion, right? I'm not exactly sure where the price escalations actually top out.
Nothing on infra costs, hardware throughput + capacity (accounting for hidden tokens) & depreciation, just a blind faith that pricing by providers "covers all costs and more". Naive estimate of 1000 tokens per search using some simplistic queries, exactly the kind of usage you don't need or want an LLM for. LLMs excel in complex queries with complex and long output. Doesn't account at all for chain-of-thought (hidden tokens) that count as output tokens by the providers but are not present in the output (surprise).
Completely skips the fact the vast majority of paid LLM users use fixed subscription pricing precisely because the API pay-per-use version would be multiples more expensive and therefore not economical.
Moving on.
That's because it's coming at things from the other end: since we can't be sure exactly what companies are doing, we're just going to look at the actual market incentives and pricing available and try to work backwards from there. And to be fair, it also cites, for instance, deepseek's paper where they talk about what their power foot margins are on inference.
> just a blind faith that pricing by providers "covers all costs and more".
It's not blind faith. I think they make a really good argument for why the pricing by providers almost certainly does cover all the costs and more. Again, including citing white papers by some of those providers.
> Naive estimate of 1000 tokens per search using some simplistic queries, exactly the kind of usage you don't need or want an LLM for.
Those token estimates were for comparing to search pricing to establish whether — relative to other things on the market — LLMs were expensive, so obviously they wanted to choose something where the domain is similar to search. That wasn't for determining whether inference was profitable or not in itself, and has absolutely no bearing on that.
> Doesn't account at all for chain-of-thought (hidden tokens) that count as output tokens by the providers but are not present in the output (surprise).
Most open-source providers provide thinking tokens in the output. Just separated by some tokens so that UI and agent software can separate it out if they want to. I believe the number of thinking tokens that Claude and GPT-5 use can be known as well: https://www.augmentcode.com/blog/developers-are-choosing-old...; typically, chain of thought tokens are also factored into API pricing in terms of what tokens you're charged for. So I have no idea what this point is supposed to mean.
> Completely skips the fact the vast majority of paid LLM users use fixed subscription pricing precisely because the API pay-per-use version would be multiples more expensive and therefore not economical.
That doesn't mean that selling inference by subscription isn't profitable either! This is a common misunderstanding of how subscriptions work. With these AI inference subscriptions, your usage is capped to ensure that the company doesn't lose too much money on you. And then the goal is with the subscriptions that most people who have a subscription will end up on average using less inference than they paid for in order to pay for those who use more so that it will equal out. And that's assuming that the upper limit on the subscription usage is actually more costly than the subscription being paid itself, and that's a pretty big assumption.
If you want something that factors in subscriptions and also does the sort of first principles analysis you want, this is a good article:
https://martinalderson.com/posts/are-openai-and-anthropic-re...
And in my opinion it seems pretty clear that basically everyone who does any kind of analysis whether black box or first principles on this comes to the conclusion that you can very easily make money on inference. The only people coming to any other conclusion are those that just look at the finances of U.S. AI companies and draw conclusions on that without doing any kind of more detailed breakdown — exactly like the article you linked me, which now I have finally been able to read, thanks to someone posting the archive link, which isn't actually making any kind of case about the subscription or unit economics of token inference whatsoever, but is instead just basing its case on the massive overinvestment of specifically open AI into gigantic hyperscale data centers, which is unrelated to the specific economics of AI itself.