Scott Alexander on AI risk: no omnipotence or hard takeoff needed, scale is the threat
slatestarcodex · x · 2026-10-07
Scott Alexander responds to @neilchilson's charge that doomers dodge the question of how intelligence becomes omnipotent in complex systems. He clarifies nobody serious expects a 12:01 AM starship — hard-takeoff believers are rare now. His actual picture: by 2035, Claude 10 in 100 million robot bodies, wired into every US biology lab, running revolutionary experiments at industrial speed. Risk scales with embedding, not omnipotence. He also points to his "Fifteen Years Of Disagreement" rebuttal of Steven Pinker's claim that doomers predict an omnipotent AI.
Related event: Scott Alexander and Neil Chilson Debate AI Doom Scenarios(4 posts)→
More from AGI Musings
- OpenAI's Navier-Stokes breakthrough shows agent-team coordination scales beyond research — cneuralnetwork · 2026-10-08
- Possible Minds May Be Narrower Than Yudkowsky Thinks, Challenging Orthogonality — jd_pressman · 2026-10-08
- Doctors Used OpenEvidence 42M Times in August, Still No Independent Evaluation — dr_alphalyrae · 2026-10-08
- Rationalist AI-apocalypse literature keeps summoning aliens, podcaster jokes — ZeroStateReflex · 2026-10-08
- Linear CEO Karri Saarinen: after tuning out AI all summer, 'nothing has really changed' — lennysan · 2026-10-08
- Yann LeCun on AI: getting the idea is easy, building it is hard, making it work is hardest — rsalakhu · 2026-10-08