The True AI Threat is People
Summary
Host Ben Bradford interviews national security expert Hamza Chaudhry about how human misuse of AI, rather than rogue superintelligence, could cause global catastrophe. They discuss risks like AI-assisted bioweapons, hallucinating nuclear defense systems, and autonomous market crashes. Chaudhry advocates for independent testing, whistleblower protections, and kill switches to prevent these "dumb dooms."
Topics discussed
Intro and technical difficulties
Introduction to AI risks and Hamza Chaudry
AI enabling bioweapons and ghost guns
Rogue states and AI-designed pandemics
Expert skepticism and cyber threats
The Stanislav Petrov nuclear false alarm story
AI in nuclear command and control systems
AI autonomy risks and reward hacking
AI-driven market crashes and cartels
Regulation, kill switches, and preemptive laws
AI benefits, conclusion, and outro
Listen ad-free on Castria