Peter Norvig on 40 years of AI: why experts get the future wrong
TheTuringPost · x · 2026-10-06
Turing Post released an 18-minute interview with Peter Norvig. His core message: even deep technical expertise doesn't protect experts from being completely wrong about AI's future.
Topics covered:
- How AI changed since the 1980s and what students still need to learn
- The return of agents, neuro-symbolic AI, whether LLMs are world models, and what today's systems still lack
- Rethinking CS education: redesigning assignments and exams, teaching students to build with AI
- AI's shift in power: can we trust AI company leaders, will AI power stay concentrated
- His views on RSI and where he agrees/disagrees with Yann LeCun
More from AGI Musings
- In the era of generating anything, what's missing is intentionality — dreamwieber · 2026-10-06
- Ethan Mollick: betting on ASI is how labs dodge hard decisions about AI today — emollick · 2026-10-06
- Timothy Gowers: we need pre-reading signals to triage AI-generated scientific papers — littmath · 2026-10-06
- Chart of the day: Agentic traffic has surpassed human traffic online for the first time — sierracatalina · 2026-10-06
- AI proof dismissed as machine slop two years ago gets an elegant human version — ctjlewis · 2026-10-06
- Every launches Thesis: 2027 conference on human work after automation, Nov 5 in Brooklyn — every · 2026-10-06