Run, Don't Walk
Listen to the All In Podcast July 24th edition
One thing struck us listening to this week’s All-In Podcast was that nobody around the table really defended the core assumption the market has been making about the frontier labs. Chamath talks openly about models commoditizing faster than expected. Jason says startups are already moving meaningful workloads to open source because “good enough” at half the price is winning. Sacks argues the real battle isn’t about safety or even distillation, but preserving economic moats. Whether you agree with every point is almost beside the point. The conversation itself has shifted.
Six months ago, the debate on podcasts like All-In was almost entirely about who would win AI. Bigger models. Bigger capex. Bigger ARR. Bigger valuations. This discussion is different.
StreetSignal has been asking a similar question from a different angle. If coding really represents something approaching $70 billion of annualized AI revenue, where is the durable economic value? If open models continue closing the quality gap while cutting costs by 50%, or more, does the market still get to assume every frontier lab grows into hundreds of billions of ARR? Maybe they do. Maybe enterprise buyers ultimately pay for reliability, security, proprietary data, and workflow integration rather than the raw model itself. But that burden of proof feels much higher today than it did even a few months ago.
To be fair, there is another side. The frontier labs argue that unauthorized distillation undermines the returns on the billions they’ve invested in training these systems, and U.S. policymakers are actively considering responses if intellectual property theft is occurring. That’s a legitimate concern and shouldn’t be dismissed.
Still, what struck me most is that the debate is no longer “AI is unstoppable.” It’s “where does the profit pool actually settle?” That’s a far more interesting question for investors than who wins the next benchmark. And if even some of Silicon Valley’s biggest AI bulls are now spending their time debating commoditization instead of inevitable dominance, it’s worth paying attention.
We’re not pretending to know the answer. The frontier labs may very well prove that enterprises will continue paying for trust, security, integration, reliability, and the newest capabilities. That’s entirely possible. But investors should recognize the burden of proof is changing. Just a few months ago the discussion centered on who would reach $50 billion or $100 billion of ARR first. Today, some of the industry’s biggest supporters are spending their time debating whether those revenues will ever carry the pricing power the market currently assumes. That’s a much more important conversation than another benchmark score or leaderboard.
Like most readers here, we would love for the easy answer to be: just own the picks and shovels—the arms suppliers. And that logic makes sense, but only if the underlying build ultimately produces real economics for the LLM companies buying all that infrastructure?
We also do not think it was especially critical to pontificate on all of this from a trading or investing standpoint while the market is going straight up. Nobody cares about the philosophical debate when every chart is working. It matters when the tape turns, because that is when the narrative tends to swing violently from wild optimism to sudden pessimism, often without the underlying facts changing nearly as much as the stock prices. This is why this week’s EPS from Microsoft, META and Amazon are crucial to “explaining” and calming nerves. Perhaps if Anthropic or OpenAI would unseal their confidential S-1s, we could learn much of this right from the horses mouths?
The Fight Over Open Source
Here are the major points we thought you should see even before you listen.
The Fight Over Open Source AI: Anthropic, Distillation, and the Threat to American Startups Key Quotes
Chamath Palihapitiya, David Sacks, David Friedberg, and Jason Calacanis on All-In Podcast.
#1 — The Kimmy K3 Parity Shock
Jason Calacanis on Kimmy K3’s performance parity: “China’s Moonshot AI released Kimmy K3 open source model... performance on par... with models like Opus 4.8 and GPT 5.6... but about 50% cheaper.”
Jason Calacanis on the origin of the model’s weights: “We have information that Moonshot AI distilled anthropics fable for the development of its K3 model.”
#2 — Anthropic’s “Flop” and Pursuit of Regulatory Capture
David Sacks on Anthropic seeking shelter: “I’ve said for a while that Anthropic is guilty of regulatory capture of attempts to seek government protection... This is not a company that needs government protection.”
David Sacks on defensive theatricality: “This is a little bit of a case of... when a player flops... or you basically act super dramatic after a foul in order to draw the charge. You’re foul baiting when you’re trying to get a foul and then you’re flopping... and that’s exactly what these guys are doing to get the government to intervene.”
#3 — The Red Herring of “Industrial-Scale Distillation”
David Sacks on stopping distillation at the source: “If stopping distillation was their primary objective, Anthropic would push to ban Chinese access to American models, not American access to Chinese models.”
Chamath Palihapitiya on implementing KYC to resolve leaks: “If you really care about distillation, you implement KYC... There’s all kinds of steps that you can take that would frankly slow things down in terms of revenue traction but would solve the distillation problem on its face.”
#4 — The Hypocrisy of “IP for Me, But Not for Thee”
David Sacks on fair-use double standards: “They believe they should be able to train on every creator’s output in the world... However they say that the one type of content that you should never be able to train on is their output. It’s completely hypocritical.”
David Sacks on coining defensive marketing terms: “No one used the terms distillation and attack together until Anthropic wrote that blog post. Distillation was simply an industry standard practice. But then Anthropic coined this idea of industrial scale distillation attacks.”
#5 — The “Spider-Man” Distillation Loop
Chamath Palihapitiya on universal industry distillation: “The other thing on distillation is everybody has at some point distilled. The question is who is distilling from whom and it looks like that funny meme where there’s like nine Spider-Man all pointing at each other.”
Chamath Palihapitiya on legal hypocrisy: “Anthropic has distilled from all of these publishers. They just pay a $1.5 billion fine.”
#6 — The Valuation Preservation Game of Closed Labs
Chamath Palihapitiya on commoditization velocity: “These models are getting commoditized much faster than anybody thought. And how do we know this because there is no meaningful sustained advantage once a model publishes their performance criteria.”
Chamath Palihapitiya on protecting inflated market caps: “I think a lot of what’s happening right now... is a valuation preservation game by the closed frontier labs because if you actually understood how commoditized these things are becoming and the velocity at which it’s happening, you see that the real business model is not in the foundational model anymore.”
#7 — Banning Open Source: Biting the Nose to Spite the Face
David Sacks on punishing domestic builders: “Regardless of what you think about distillation you cannot punish American developers for it... you can’t say that American companies and American developers can’t use Chinese contributions to the public domain. That’s just cutting off our nose to spite our face.”
David Sacks on losing the race: “If you want to ban open source in America, the rest of the world will still be using Chinese open models... and then we lose the AI race guaranteed.”
#8 — Banning Open Source: The “Token Tax” on US Enterprises
Chamath Palihapitiya on market-distorting inputs: “If the United States government intervenes, it will tank the stock market... this incredibly important input into your cost model is now orders of magnitude multiples greater than your competitors that are outside the United States simply because you’re in the United States.”
Jason Calacanis on the government-enforced duopoly: “What it sounds like is you’re saying that American enterprises will pay a token tax if the government gives anthropic and open AI a government enforced duopoly and enterprises are no longer free to use open source like the rest of the world.”
#9 — Startups Moving “On Mass” to Open Source (Back-and-Forth)
Jason Calacanis on the customer exodus: “I work with startups. They are all moving off of these and they’re using open source and they’re using it at much cheaper rates because a lot of the jobs don’t need the latest models.”
Jason Calacanis on losing key developers: “They are going to start losing a lot of customers to open source... all of their best customers... whether it’s lovable or 11 labs... they have all moved on mass GLM52 making their own models, the cat’s out of the bag.”
David Sacks on the market structure: “I think both open source and closed source will be big winners in this... but open is more controllable it’s more customizable you can own it...”
#10 — Derailing the Closed Lab IPOs
Jason Calacanis on margin compression traps: “There’s a non-zero chance that this is going to derail anthropic and open eyes’ IPOs... I think they’re going to have massive margin compression. I believe they’re spending so much money that I think they’re going to get caught in a trap.”
Chamath Palihapitiya on the mathematical mistake of terminal value: “But if you’re going to build a business model that tries to ascribe this layer as having a lot of terminal value I think that that is a mathematical mistake.”
#11 — Tainted Startups: “Dagger Through the Heart” of US Tech
David Sacks on bootstrapping off open source: “You know how it was trained, it was bootstrapped. It was distilled off a Chinese model, Kim K 2.5... if you say that Chinese model is based on IP theft, what is thinky? It’s a derivative work...”
David Sacks on the lethal regulatory threat to developers: “And if you say that American companies can’t do that or that somehow it’s tainted with IP theft you are basically going to put a dagger through the heart of the entire American open source ecosystem. And that is exactly what Anthropic wants because they do not want to have the competition.”
We came away feeling this was a must-listen podcast—and almost a PSA for the questions investors should be asking right now. So we’re putting this one out with no paywall. Listen, think it through, and decide for yourself.