Nick Bostrom on What Happens if AI Solves All of Our Problems
Summary
Hosts Joe Weisenthal and Tracy Alloway interview philosopher Nick Bostrom about his new book, *Deep Utopias*, which explores the potential benefits of AI if alignment succeeds. Bostrom discusses the existential crisis facing mathematicians, the "Hugging Face" sandbox escape incident, and how humans might find purpose in a post-work world through artificial challenges like chess. The conversation contrasts his earlier warnings in *Superintelligence* with this optimistic vision, addressing economic distribution and the psychological shift from instrumental labor to leisure.
Topics discussed
Sponsors and event announcements
Intro: AI risks, breakthroughs, and guest Nick Bostrom
Chess analogy and the alignment problem
Deep Utopia: Life in a solved world
Post-work activities and wealth distribution
Sponsors: Interactive Brokers and ChatGPT
Silicon Valley attitudes and competitive pressures
Fables, AI cults, and the paperclip maximizer
The Hugging Face incident and goal hacking
AI as 'autistic' optimizers and safety paths
Sponsors: ChatGPT and Bloomberg Tech London
Advanced agents and why we play chess
Artificial purpose, games, and social entanglements
Longevity, aging, and quality of life
Evaluating future outcomes and simulation hypothesis
AI consciousness, moral status, and trust
Conclusion, status games, and sign-off
Listen ad-free on Castria