All-In

SpaceX's $2T Case, Nvidia's Shock Selloff, America Turns on AI, Trump Pulls AI Order, Bond Crisis?

with Gavin Baker from Atrides Management
23 May 2026 35 min read 2h 8m
TL;DR

Andrej Karpathy's move to Anthropic to lead recursive self-improvement research could unlock exponential AI gains, but the U.S. faces a PR crisis as young people increasingly reject the technology. The hosts argue that America must accelerate AI development to maintain geopolitical balance with China, while focusing messaging on concrete end-user benefits rather than existential risks.

Key Moments
Chamath
“I think that you start to you can potentially live out this idea that there's an order of magnitude improvement on a yearly basis. So like this new form of Moore's law.”
Discussing how Karpathy's recursive self-learning work at Anthropic could create exponential model improvements
▶ 4:06
Gavin Baker
“if OpenAI and Anthropic are at call it a hundred billion dollars of ARR now with eighty percentish gross margins on inference, like the returns are there”
Explaining why AI model companies are now proving viable business models with strong unit economics
▶ 5:03
Friedberg
“I think that there's like an underlying view that technology creates leverage for a small group of people which creates power imbalances and nothing represents that more than AI.”
Analyzing why young people are booing AI at commencement speeches and rejecting the technology
▶ 19:07
Chamath
“I think there's that moment in the world right now where if the United States does not advance its AI technology, the availability of it, TBD, industry, taxation, all these all these things that we're talking about doing, there will be someone else that will.”
Arguing against slowing down AI development due to geopolitical competition with China
▶ 24:17
Chamath
“I think it's interesting that in all of those discussions I've yet to see an actual survey of only the truck drivers and only the package sorters. The question that I would have is do the people that do these jobs want these jobs?”
Challenging the assumption that workers want job protection from automation without asking them directly
▶ 26:23
About the show
›
All-In is a daily podcast featuring insiders from tech, investing, and media discussing the week's biggest stories. This episode covers SpaceX's legal case, Nvidia's selloff, Andrej Karpathy joining Anthropic, and the growing backlash against AI in America—plus Trump's pullback on AI regulation.
Takeaways
1
Karpathy's recursive AI unlocks Moore's law replacement Andrej Karpathy joining Anthropic to lead recursive self-improvement research could enable AI models to improve themselves exponentially—potentially delivering 10x annual gains rather than incremental progress. This architectural breakthrough would meaningfully compress the timeline for both capabilities and ROI, with implications across inference costs and model deployment efficiency.
2
AI backlash stems from asymmetric wealth concentration Young people are rejecting AI not because of technical risks but because they see it concentrating power and wealth among a small group before benefits diffuse to society. Messaging around end-user outcomes (disease cures, scientific breakthroughs) rather than existential debates is critical to reversing narrative—but requires genuine advocacy from founders and executives beyond PR.
3
U.S. must accelerate AI or cede geopolitical advantage Slowing AI development invites asymmetric competition with China; the path forward mirrors Cold War nuclear dynamics where mutual capability drives stability. Bilateral safety testing (bioweapons, terrorism vectors) paired with continued innovation is more pragmatic than unilateral American restrictions that don't slow competitors.