Stopping The AI Safety Cult ft. Adam Becker & Cal Newport
Summary
Cal Newport and Adam Becker explore the origins of the rationalist and effective altruist movements, tracing their roots to 1990s transhumanist forums and Eliezer Yudkowsky’s early writings. They detail how these communities, fueled by Peter Thiel’s funding, deeply influenced the founding and leadership of major AI labs like OpenAI, Anthropic, and DeepMind. The episode argues that this eschatological belief in an imminent AI singularity drives reckless safety practices and apocalyptic rhetoric within the industry. Ultimately, the hosts contend that this unique blend of futurist philosophy a…
Topics discussed
Sponsorships: ZipBound, Dupixent, and M&M's
Sponsorships: Mastocytosis story and OxyClean
Introduction and the AI existential risk debate
Origins of AI doom in 90s transhumanist forums
Nick Bostrom and the shift to AI safety concerns
Yudkowsky, LessWrong, and the rise of rationalism
The Harry Potter fanfic and early EA funding
Yudkowsky's shift from excitement to fear
Sponsorships: VPN, ZipBound, and Prime Video
DeepMind founding and Peter Thiel's influence
Defining 'alignment' and the OpenAI board split
Critique of EA actions and longtermism
EA recruitment tactics and prediction markets
The mix of money, tech, and apocalyptic views
OpenAI's negligent agent experiments
Sponsorships: Surfshark, ZipBound, and Pepperidge
Jacob Coxon's resignation and industry response
Analyzing the EA schism and coordinated leaks
Jensen Huang's reaction and China's AI approach
Advice for navigating AI narratives and terminology
Conclusion and final sponsorships
Listen ad-free on Castria