toly 🇺🇸

Boooooo

Holy smokes

Blessed are the defenders

@SimkinStepan Yet you can’t beat the spy etf

My biggest takeaway from this World Cup is that Lenovo has an AI laptop and that Motorola somehow is still around

Blessed are the goalkeepers

VAMOOOOOOOSSSSS

@RealSpaceMonkey 🐐🐐🐐

Gonna make a prediction without knowing anything about football and say Argentina will win against Spain despite what the market thinks

my quantum

@MikeBradleyAI Just like they do for email

**Also just for what it’s worth here all jokes aside. This is already happening. I was hanging out with a guy the other day who manages the inference stack and hardware for a major hospital system. If you think nobody is going to deploy in house hardware at scale I promise you they will and already are.

This is exactly the state of things BTW. Individuals, enterprises, companies, hospitals, all being able to host frontier AI themselves destroys the business models of these large frontier labs. They are hoping for 1T USD IPO’s they aren’t going to sit on their hands while their moat evaporates. But they also aren’t managing to compete their way into staying ahead. All that’s left is attempts at PR and regulation and pushing a “free and open equals bad” narrative that couldn’t be more transparently self serving.

LFG champions of free markets 🇦🇷🇦🇷🇦🇷

Imagine building on legacy BaaS APIs at a time when financial infrastructure is becoming truly programmable while agents are becoming capable of autonomously operating it

@matteodotsui Gotta root for the last bastion of capitalism

@toly Who you got winning Toly?

2 yellow cards each

@grubbyhat @RickBakas youtu.be/l8ukak8P2vY?is…

@toly @RickBakas You cant be coerced by a media company? Rupert Murdoch would disagree

The irony is that they are fixing everything. A farm whose workers devour more food than the farm yields will starve. Feeding the hungry requires the farm the produce more food then the workers consume. That’s the profit. Lifting people out of poverty can only come from the profitable work of someone else. Comrades, the emancipation of the proletariat can only be achieved by a massive increase of productive forces. Productivity improvements are risky, risk requires capital. 99% of what Elon and Bezos do is underwrite risk with capital to improve future productivity.

@RickBakas Government is what does the coercion. I literally built a phone to compete with Google and Apple and had ~100m in sales.

They can’t coerce by owning a media company. I am a voluntary user of X. Only the government can coerce. I am an involuntary payer of taxes. You aren’t influenced by X otherwise you’d be an Elon fan boy. So somehow you resisted his coercion. But you personally believe that the unwashed prolles all must be coerced by his propaganda and need the firm hand of the intelligencia to guide them.

@AISystemGuy @ylecun @hosseeb @deanwball That’s awesome!

@toly @ylecun @hosseeb @deanwball For context: my current revenue comes from selling SFT/DPO models, so I pay close attention to model licensing terms. permissionless-labs.com/services.html

This argument by @deanwball is being badly misunderstood. It's OK to disagree with it, but first you have to actually understand what he's saying. He's saying: releasing the weights for a frontier-level model is effectively dumping. Dumping is when you sell a product at significantly below cost in order to corner market share. It's illegal. The reason: dumping results in short-term consumer surplus, but long-term it prevents the formation of a competitive market and discourages capex outside of the dumper. Standard Oil famously did this in order to consolidate the oil market before it was broken up. So why is he claiming releasing the weights of a frontier level model is basically dumping? Isn't he just describing open source? His argument: it's not financially sustainable to train a frontier model and release the weights. In the long run, you will not be able to internalize enough of the gains given the cost of training a frontier model, because neoclouds and other inference providers will be able to outcompete you at actually serving the model. It costs an astronomical amount of money to train frontier models, and if everyone else can serve them, you don't capture enough of the surplus to pay for the training and R&D. It's not like normal open source when you build some software and then release it and sell services on top of it. The amount of capex required for frontier-level models is an order of magnitude higher than normal software, which is why doing this at frontier level is so economically irrational. Right now the Hong Kong stock market is ebullient enough that Chinese AI companies are not getting punished for the fact that they're all deeply, deeply unprofitable. Releasing model weights is great marketing, intellectually appealing, and strikes fear into the hearts of their opponents. We can assume the status quo continues for a while because of the AI supercycle. But eventually the AI market will correct, the Hong Kong market will dump, and suddenly these Chinese labs won't be able to afford to training super expensive models without internalizing more of the gains. But what if China, seeing that this strategy is successfully kneecapping the US lead (by discouraging further capex and lowering valuations), says no--don't stop. And so the Chinese government starts buying up the shares of these companies and demanding that they continue releasing frontier-level weights, profitable or not. In that case, it becomes a genuine space race. For-profit companies cannot continue to compete on either side. US labs valuations fall, and the White House realizes that to keep their advantage in the AI race, they cannot rely on the free market to maintain their lead. They nationalize the labs and fund them off government subsidies. Now you have government-controlled and distributed models on both sides. That's what Dean is calling the "dystopian hellscape." The best analogy is drug development: if China were to sell American drugs back to us really cheaply, that would result in a large short-term consumer surplus. Cheap Viagra and Ozempic is obviously great. But in the long run, this would discourage investment in developing new drugs. That's the sense that Dean is saying it's long-term "decel." Now, I happen to disagree with Dean. I think the consumer surplus of having frontier-level open weight models is huge, even at the current capabilities. I also think China is going to defect from this strategy soon (there's been reporting along these lines, that Beijing will stop allowing large models to be open-weight; I think there are other reasons for this aside from competition). I also suspect that nationalization of labs is inevitable as they take on more geopolitical and cyber capabilities. But he's not wrong--releasing frontier-level weight models is weird. The question of how long this market will remain profit-driven is a very coherent question to ask.

@AISystemGuy @ylecun @hosseeb @deanwball Yea. Thats totally fine.

That's a licensing question. Distillation is a standard ML technique. Whether you can distill from a particular commercial model depends on the provider's terms. For example, Anthropic says using Claude to distill a competing general-purpose LLM violates its terms. Using Claude to generate data for a narrow, non-competing model (e.g. a sentiment classifier or document classifier) is generally permitted. Technically, distillation has been a standard ML method for over a decade: arxiv.org/abs/1503.02531