Gavin Baker

1 is just super obviously not true. If there was one lab, and only one lab, with 99.999999% margins, that would be a disaster for the neoclouds and everyone who supplied that single lab. How can this not be obvious to you? There are somewhere between 1000 and 1 million case studies across a variety of industries showing that monopsonies exercise power over their suppliers. Fragmentation and competition at the model layer is good for everyone who is not at the model layer. At an extreme - a world where companies like Latham & Watkins run fine-tuned open-source models on their own GPUs and Nvidia funds the open-source training runs - it is possible for there to be zero profits at the model layer and yet high returns for AI infrastructure supplies (semiconductors, datacenters and power in this example). This is all really basic stuff. You seem smart. Think for maybe 5-10 seconds before you post a reply.

A better analogy would be: what happened to the consumption of steel and oil after the automobile entered mass production? And yes, steel and oil were commodities but not all of the inputs into producing tokens are commodities. The valuations of Coreweave, Oracle and Nvidia should logically be negatively correlated with margins at the model layer especially at an extreme. The higher the margin added at the model layer on top of the cost of producing a token (with the cost of producing a token effectively being Coreweave, Oracle and Nvidia’s revenue), then less tokens will be produced all else equal and therefore less revenue will be generated by Coreweave, Oracle and Nvidia. Maybe it would help your understanding to invert this. It would obviously be negative for everyone in AI infrastructure (inclusive of your Coreweave, Oracle and Nvidia) if OpenAI and Anthropic were a monopsony buyer of compute at the model layer. And terrible for the world if they were a monopoly seller of AI. And I wish I had the luxury of engaging in spin. I am accountable for my hypotheses about the future in the realest way.

Wild 24 hours for AI and lots of different proposals have been made. TLDR; the only *tangible* new fact is that OpenAI and Anthropic are going to have embedded 3rd party evaluators from unknown organizations with Dario floating METR as a possibility. Having 3rd party evaluators is smart as there is no Section 230 style liability shield for model outputs and showing a “duty of care” will be important in future litigation. Several internet companies might have gone bankrupt without Section 230 so limiting liability really matters. There are minimal investment implications from this single new fact, but I do think that for anyone who wants a “smoother for longer” cycle then most constraints are good: wafers, watts, real rates and spreads. Excessive regulation is a different matter but I don’t think we are anywhere close to this even if the vector changed over the last 24 hours. To summarize the events: Dario made the most maximalist proposal of the weekend: embedded 3rd party evaluators, a national regulatory regime for models beyond a certain capability/ingredient threshold, a broad international regulatory pact between democracies, stricter limits on compute/distillation for China and then a different international regulatory regime that encompasses China. Before there is a national regulatory regime, he wants a Sherman act waiver so that Anthropic can safely coordinate with OpenAI and other frontier labs without antitrust fears. TBF, this latest proposal is much less maximalist than some of his prior proposals like “Policy on the AI Exponential,” where he advocated for an FAA for AI. I believe he is sincere in his beliefs. And despite all the protestations, all of this would also probably be good for his business over the long-term. Sam agreed that embedded 3rd party evaluators were a good idea and stated they would implement them. Again, this is smart as should help limit future liability. Elon said “Dario is right” and later specified that “Dario is right that there should be some oversight. Peer review of AI by competitors is the right way to start this off.” This would be a MPAA like self-regulatory structure for AI with regular calls between the labs plus a process where each new model is evaluated for safety by competitors for a 1-2 week period before being released. That is *wildly* different from Dario’s proposal and in-line with what David Sacks has been proposing. Elon also stated that nothing was going to slow down open-weight models. Demis said that Dario’s essay was a “step in the right direction.” Dario also said that he was also open to Demis’ idea of a FINRA like self-regulatory structure as part of his proposal. David Sacks had a thoughtful post where he said that Dario and Sam should pace unilaterally, called the antitrust waiver a cartel request and denied that METR was truly independent given their ties to Anthropic. Sriram Krishnan, former White House AI advisor, noted that it would be important to have the 3rd party evaluators come from independent organizations that are not affiliated with any lab, which is basically an indirect statement about the relationship between METR and Anthropic which Sacks was explicit about. Clem from Hugging Face said they were open to being a neutral 3rd party evaluator, which is interesting especially if Jensen was consulted before that post. Alexander Wang from Meta noted that alignment would be an increasing focus going forward. An executive order seems likely after all this and the language in this EO is going to be really important. It is possible to democratize and distribute AI broadly and safely without centralizing it in the hands of a few corporations who might each become more powerful than any single government. I do not want a few humans in control of intelligence. I want us all to have our own intelligences that reflect our own values and human variation in all of its richness. Intelligence distribution over intelligence centralization FTW.

Anthropic pre-IPO gamesmanship post. Pure speculation but sharing as curious for thoughts. Anthropic shifted from gross to net ARR accounting and stripped out both Meta and Chinese distillation from their $65 billion ARR number. Meta speculated to be over $5 billion in ARR so taking them out means they can easily weather it when Meta turns them off shortly after being public, which is widely expected. Also decreases the odds of Meta turning them off, watermelon quality dependent. All smart. Then release Fable 5.1 so OpenAI feels confident releasing Astra. Vibes here on Astra are really good btw. I think that Astra was probably better than Anthropic was expecting. Now there are whispers that Anthropic has solved Navier-Stokes, which would be super impressive. Anthropic probably releases Fable 5.2, which should be better than Astra unless something is awry, sometime before the IPO. Likely also planning on showing a significant reacceleration in ARR in September which will of course leak to the press. Grok 4.7, Meta’s Watermelon and ChatGPT 6.1 all likely coming in the next 6 weeks as well. All those labs are confident about their roadmaps in a way I have not seen in the last 18 months. And we will see about Gemini 4. Competitors get a vote in all these plans. Grok Bot feels like the best agentic harness yet for enterprise use cases and Instinct is a promising agentic harness for consumer use cases. Should see variations of both from competitors soon. Grok Bot remains transformational for my use cases. And all this is happening into a continued acceleration in overall AI demand. Wild times. As an aside, I think Krishna might turn out to be an exceptional CFO. His former Blackstone colleagues speak super highly of him. Going to be important as communicating clearly to Wall Street if they decide to shift their compute from inference to training will be difficult to digest the first time. Probably worth studying Amazon’s invest and then “check-in” margin strategy from 2010 through maybe 2016, which investors eventually understood.

@corleonecapital What do you think Astra is? And Grok 4.6? And Muse 1.3? And Blackwell was almost certainly used for Fable 5.1 post-training.

The Hugging Face acquisition is important for America and I think Nvidia will be a good steward for the ecosystem. The Poolside transaction may end up mattering even more. I think Jensen is likely to bring American open-weight AI to the frontier, which is going to be awesome for America. We might see a multi-billion $ training run from Nvidia in the next 18 months for Nemotron v5-6 that is easy for customers to post-train and optimize for their own use case. Would be cool to see a 10 trillion plus parameter American open-weight model. If the best open-weight base is American, cheap to run, and actually post-trainable, then people, companies, labs, and governments can own their own intelligence instead of renting it from a company that might not share their values. Would be good for freedom to have a rich variety of AIs that reflect our own individual human preferences. And for the sake of the clarity, I think cheaper, specialized open-weight intelligence might end up making frontier tokens more valuable!

The Hugging Face acquisition is important for America and I think Nvidia will be a good steward for the ecosystem. The Poolside transaction may end up mattering even more. I think Jensen is likely to bring American open-weight AI to the frontier, which is going to be awesome for America. We might see a multi-billion $ training run from Nvidia in the next 18 months for Nemotron v5-6 that is easy for customers to post-train and optimize for their own use case. Would be cool to see a 10 trillion plus parameter American open-weight model. If the best open-weight base is American, cheap to run, and actually post-trainable, then people, companies, labs, and governments can own their own intelligence instead of renting it from a company that might not share their values. Would be good for freedom to have a rich variety of AIs that reflect our own individual human preferences. And for the sake of the clarity, I think cheaper, specialized open-weight intelligence might end up making frontier tokens more valuable!

OpenAI still taking share while Anthropic likely reaccelerating. And open source is growing even faster. https://t.co/CeXwcseNAH

Should note the real comparison will be with Rubin Ultra given Rubin shipping today so kinda have to wait. And Feynmann coming fast. But still impressive debut from Jalapeño.

Impressive that Jalapeño outperforms the comparable TPU and is in the mix with Rubin. Credit where credit is due - first good ASIC outside of TPU/Trainium. However, will likely significantly underperform a disaggregated GPU/Trainum plus SRAM accelerator setup. Especially with AFFN disaggregation.