80,000 Hours Podcast What everyone is missing about Anthropic vs the Pentagon. And: The Meta leaks are worse than you think.
150 snips
Apr 3, 2026 A deep look at the clash between a major AI firm and the Pentagon, weighing claims of hypocrisy, naivety, and anti-democratic behavior. A shocking dive into leaked Meta documents reveals massive scam ad revenue and how internal anti-fraud fixes were sidelined. Final segments sketch policy ideas, including stricter oversight for AI and tougher penalties for platforms profiting from scams.
AI Snips
Chapters
Transcript
Episode notes
Opposing A Specific Government Move Is Not Hypocrisy
- Opposing a specific government action doesn't mean opposing all AI regulation.
- Rob Wiblin argues that who controls AI matters as much as whether government controls it, so objecting to a specific coercive move is consistent with supporting oversight.
Legal Pushback Can Protect The Whole AI Industry
- Resisting government pressure can create protective legal precedents for the whole industry.
- Rob notes Anthropic's stand rallied competitors and legal analysts expect a good chance of court success, potentially blocking economic coercion.
Power Does Not Make Actions Right
- Powerful-actor inevitability is a poor justification for accepting harmful actions.
- Rob critiques Ben Thompson's realist claim that state power makes actions acceptable, distinguishing prediction from moral acceptance.