What If Artificial Intelligence Progress Explodes? (with Benjamin Todd)

Benjamin Todd, co-founder of 80,000 Hours, joins Dan and Henry to discuss whether artificial intelligence progress could become explosive.Benjamin explains why he thinks transformative artificial intelligence by 2030 is a serious possibility, how feedback loops in artificial intelligence research could accelerate progress, and why the most important risks now go beyond classic alignment problems. The conversation covers artificial intelligence timelines, bottlenecks in chips and research talent, the future of work, mass unemployment, concentration of power, engineered pandemics, space governance, and how young people should think about their careers in a rapidly changing world.Topics discussed include:• Why 80,000 Hours increasingly focuses on artificial intelligence• The case for short timelines to transformative artificial intelligence• Whether artificial intelligence progress could become explosive• Feedback loops in artificial intelligence research• Chip bottlenecks, data centres, and geopolitical risk• Whether artificial intelligence will cause mass unemployment• Why “become a plumber” may be bad career advice• Alignment, control, and concentration of power• Misuse risks, engineered pandemics, and future governance• How to think clearly under extreme uncertaintyBenjamin Todd is the co-founder of 80,000 Hours and the author of 80,000 Hours, a new book about how to choose a career that is both personally rewarding and socially impactful. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.conspicuouscognition.com/subscribe