Yampolskiy: Superintelligence needs no body to wipe out 99% of humanity
romanyam · x · 2026-09-12
AI safety researcher Roman Yampolskiy argues superintelligence could wipe out 99% of humanity — just not Terminator-style:
- His core analogy: ask a squirrel how a human would kill it, and it says "chase me, climb a tree, bite me hard" — useless for grasping the real threat. Likewise, humans cannot predict what the smartest system will do.
- A superintelligence would develop new technology and discover new laws of physics; you won't know what gets you or how.
- Responding to "it doesn't even have hands" objections, he insists that demanding a specific mechanism misunderstands what superintelligence is.
More from AGI Musings
- Bengio lays out analysis of misaligned agents; sebkrier pitches adversarial multi-agent checks — sebkrier · 2026-09-12
- Doomer take: mathematicians have one year before open AI models sweep Millennium Prizes — teortaxesTex · 2026-09-12
- The Real AI Doom Scenario Isn't Terminator, It's a Mental WALL-E — eigenhector · 2026-09-12
- Maybe Cloud Atlas Got the AI Apocalypse Right, Says Thread — eigenhector · 2026-09-12
- The final UI is clear: language in, intelligence, language or generated UI out — signulll · 2026-09-12
- AI Doom Was Never Terminator, Just People Outsourcing Thinking to Robots — eigenhector · 2026-09-12