6 ms·
ICYMI, Amodei said the same in much greater detail: "If you consider each model to be a company, the model that was trained in 2023 was profitable. You paid $1
by dcre 1y ago
ICYMI, Amodei said the same in much greater detail:
"If you consider each model to be a company, the model that was trained in 2023 was profitable. You paid $100 million, and then it made $200 million of revenue. There's some cost to inference with the model, but let's just assume, in this cartoonish cartoon example, that even if you add those two up, you're kind of in a good state. So, if every model was a company, the model, in this example, is actually profitable.
What's going on is that at the same time as you're reaping the benefits from one company, you're founding another company that's much more expensive and requires much more upfront R&D investment. And so the way that it's going to shake out is this will keep going up until the numbers go very large and the models can't get larger, and then it'll be a large, very profitable business, or, at some point, the models will stop getting better, right? The march to AGI will be halted for some reason, and then perhaps it'll be some overhang. So, there'll be a one-time, 'Oh man, we spent a lot of money and we didn't get anything for it.' And then the business returns to whatever scale it was at."
https://cheekypint.substack.com/p/a-cheeky-pint-with-anthropic-ceo https://cheekypint.substack.com/p/a-cheeky-pint-with-anthrop...
- selimthegrim 1y agoThis sounds like fabs.
- whatshisface 1y agoI don't see why the declining marginal returns can't be continuous.
- 827a 1y agoOpenAI and Anthropic have very different customer bases and usage profiles. I'd estimate a significantly higher percentage of Anthropic's tokens are paid by the customer than OpenAI's. The ChatGPT free tier is magnitudes more popular than Claude's free tier, and Anthropic in all likelihood does a higher percentage of API business versus consumer business than OpenAI does. In other words, its possible this story is correct and true for Anthropic, but not true for OpenAI.
- dcre 1y agoGood point, very possible that Altman is excluding free tier as a marketing cost even if it loses more than they make on paid customers. On the other hand they may be able to cut free tier costs a lot by having the model router send queries to gpt-5-mini where before they were going to 4o.
- DenisM 1y agoFree tier provides a lot of training material. Every time you correct ChatGPT on its mistakes you’re giving them knowledge that’s not in any book or website. Thats a moat, albeit one that is slow to build.
- jacurtis 1y agoThis is very true. ChatGPT has a very generous free tier. I used to pay for it, but realized I was never really hitting the limits of what is needed to pay for it. However, at the same time, I was using Claude much less, really preferring the answers from it most of the time, and constantly being hit with limits. So guess what I did. I cancelled my OpenAI subscription and moved to Anthropic. Not only do i get Claude Code, which OpenAI really has no serious competitor for. I still use both models but never run into problems with OpenAI, so i see no reason to pay for it.
- pera 1y agoCopy laundering as a service is only profitable when you discount future settlements: https://www.reuters.com/legal/government/anthropics-surprise-settlement-adds-new-wrinkle-ai-copyright-war-2025-08-27/ https://www.reuters.com/legal/government/anthropics-surprise...
- meshugaas 1y agoThe "model as company" metaphor makes no sense. It should actually be models are products, like a shoe. Nike spends money developing a shoe, then building it, then they sell it, and ideally those R&D costs are made up in shoe sales. But you still have to run the whole company outside of that. Also, in Nike's case, as they grow they get better at making more shoes for cheaper. LLM model providers tell us that every new model (shoe) costs multiples more than the last one to develop. If they make 2x revenue on training, like he's said, to be profitable they have to either double prices or double users every year, or stop making new models.
- renjimen 1y agoBut new models to date have cost more than the previous ones to create, often by an order of magnitude, so the shoe metaphor falls apart. A better metaphor would be oil and gas production, where existing oil and gas fields are either already finished (i.e. model is no longer SOTA -- no longer making a return on investment) or currently producing (SOTA inference -- making a return on investment). The key similarity with AI is new oil and gas fields are increasingly expensive to bring online because they are harder to make economical than the first ones we stumbled across bubbling up in the desert, and that's even with technological innovation. That is to say, the low hanging fruit is long gone.
- meshugaas 1y agoexactly: it’s like making shoes if you’re really bad at making shoes :)
- runako 1y ago> new models to date have cost more than the previous ones to create This largely was the case in software in the '80s-'10s (when versions largely disappeared) and still is the case in hardware. iPhone 17 will certainly cost far more to develop than did iPhone 10 or 5. iPhone 5 cost far more than 3G, etc.
- Romario77 1y ago
- LZ_Khan 1y agoAlso Amodei has an assumption that a 100m model will make 200m of revenue but a 1B model will make 2B of revenue. Does that really hold up? There's no phenomenon that prevents them from only making 200m of revenue off a $1B model.
- Avshalom 1y agoOkay but noticeably he invents two numbers then pretends that a third number is irrelevant in order to claim that each model (which is not a company) is a profitable company. You'd think maybe the CEO might be able to give a ball park on the profit made off that 2023 model. ETA: "You paid $100 million... There's some cost to inference with the model, but let's just assume ... that even if you add those two up, you're kind of in a good state." You see this right? He literally says that if you assume revenue exceeds costs then it's profitable. He doesn't actually say that it does though.
- DenisM 1y agoFantastic perspective. Basically each new company puts competitive pressure on the previous company, and together they compress margins. They are racing themselves to the bottom. I imagine they know this and bet on AGI primacy.
- oblio 1y ago> I imagine they know this and bet on AGI primacy. Just like Uber and Tesla are betting on self driving cars. I think it's been 10 years now ("any minute now").
- kgwgk 1y ago>> If we didn't pay for training, we'd be a very profitable company. > ICYMI, Amodei said the same No. He says that even paying for training a model is profitable. It makes more revenue that it costs - all things considered. A much stronger claim.
- dcre 1y agoI take them to be saying the same thing — the difference is that Altman is referring to the training of the next model happening now, while Amodei is referring to the training months ago of the model you're currently earning money back on through inference.
- kgwgk 1y agoMaybe he means that but the quote says “We're profitable on inference.” - not “We're profitable on inference including training of that model.”