Quote · All-In with Chamath, Jason, Sacks & Friedberg
Open Source Wins, AGI Is Here, and Scorsese's AI Toolkit with CEOs of Cerebras & Black Forest Labs
Where this was said
Open Source, AI Sovereignty, and the Road to AGI
At 30:49 · chapter starts 16:28
Feldman takes the conversation into longer time horizons. The fundamental constraint on human knowledge is that paradigms shift at the speed of generations: Thomas Kuhn showed us that Freud, Skinner, and their disciples maintained intellectual dominance not because their ideas were right, but because they held positions of leadership until they died. New ideas have to wait for the old guard to exit. AI breaks this constraint. Feldman reaches for the geneticist's tool: fruit flies produce two generations per day, allowing researchers to observe thousands of generations of evolution in a single career. AI is now doing the equivalent for knowledge — compressing what previously took centuries of human intellectual turnover into near-instantaneous iterative cycles. He closes with the Palace of Versailles as metaphor: the builders who worked on century-long projects across multiple family generations were doing recursive, compounding learning. AI is that, but at the speed of compute.
The open source AI landscape has quietly become a geopolitical flashpoint. Outside of OpenAI's OSS model, most available open source options are Chinese. Regulated industries in finance and healthcare that need on-premise, sovereignty-friendly AI have almost nowhere to turn for domestic alternatives.
Nikesh Arora of Palo Alto Networks told Jason Calacanis that testing a frontier AI model against their own systems was devastating — it found critical bugs in hours that their own team had missed. They halted everything for six weeks of emergency patching. That's the case for government red teaming in a single anecdote.
Nikesh Arora of Palo Alto Networks told Jason Calacanis that when tested with a frontier AI model, it found unknown critical bugs — forcing the company to halt operations and patch for six weeks.
By any definition anyone would have offered 10, 20, or 50 years ago — including the Turing test — we have already hit AGI. The goalposts moved because the reality arrived faster than our imagination could keep up.
Feldman argues that by any definition of AGI used 10, 20, or 50 years ago — including the Turing test — AI has already blown past it.
