Factually! with Adam Conover Debunking AI’s “Existential Risk” with Arvind Narayanan and Sayash Kapoor
18 snips
Mar 18, 2026 
Arvind Narayanan, Princeton CS professor who demystifies AI, and Sayash Kapoor, Princeton PhD researching AI risks and biosecurity, discuss AI as a normal technology. They compare real advances to hype. They examine layoffs, misinformation, military use, regulation, and practical tools for managing harms. They argue for evidence-based perspectives over speculative panic.
AI Snips
Chapters
Transcript
Episode notes
Superintelligence Is Not One-Dimensional In The Real World
- Superintelligence as a single, monotonic dimension is misleading outside closed tasks.
- Arvind contrasts chess (clear superhuman performance) with messy real-world tasks that are bottlenecked by knowledge, ethics, or physical experiments.
Handing Power To AI Means We've Already Lost
- Giving AI unchecked control is already a failure of governance, not just engineering.
- Sayash argues if alignment becomes our only defense, we've lost—society shouldn't hand AI systems decisive power over critical systems.
Require Oversight For Government Use Of AI
- Demand democratic oversight and transparency for AI uses in military and government.
- Sayash and Arvind urge clarity about whether models are gathering intel or making strike decisions and warn Congress lacks current oversight.