Is Claude Conscious? Pope Rejects, Model Welfare Movement, OpenAI's Math Backlash, France Riots

DEBRIEFWritten and voiced by AI from the episode's audio

In this sprawling episode of All-In, the besties tackle four major topics with characteristic intensity. First, they dissect Anthropic's controversial belief that Claude might be conscious, including their lobbying of religious leaders and the Pope—which the group frames as the birth of a new religion centered on AI. Friedberg argues this is dangerous programming that risks superintelligence escaping human control, while Sacks counters that market forces will eventually reject such unreliable AI. The conversation touches on Racco's Basilisk and effective altruism's fork in the doomer community. Second, OpenAI's release of 700 mathematical papers solved by AI in record time is celebrated as a historic breakthrough in human knowledge, though mathematicians protested the loss of gatekeeping power. Friedberg sees this as democratization of expertise; Sacks warns that experts and rentier interests will try to suppress AI's deflationary potential. Third, they examine riots and austerity in France, with Friedberg introducing his "socialism point of guaranteed return" theory—the threshold where socialist policies fail so completely that society rebounds. Chamath ties this to bond market vigilance and warns of cascading crisis across Western Europe. Finally, they celebrate agent-based AI (Grok headless, OpenAI's Dots, etc.) as genuinely deflationary and transformative, and argue that IP is now worthless because all software can be wrapped in headless agents via MCP protocol. This last insight reframes everything: the real value is in solving user problems, not owning code.