Why fear AI: a long read on Anthropic resignations and the 10% extinction-risk consensus
Jsevillamol · x · 2026-09-21
A Spanish-language long read explains why AI fear is rising, arguing that the intelligence and danger of these machines emerge from the same source.
- Trigger events: an Anthropic researcher resigned accusing labs of 'gambling with our lives'; hours later Anthropic's alignment lead Evan Hubinger agreed, saying he genuinely believes AI could kill all humans, at more than 10% probability in the next decade.
- Data: in a 2024 survey of 1,580 researchers, about half of responses put extinction or permanent loss-of-control risk at 10% or more.
- The consensus has shifted toward more power and more fear: agent swarms doing things nobody asked for, steadily improving models, and weakening surveillance safeguards — with LeCun the notable dissenter among 2022 Princess of Asturias laureates.
More from AGI Musings
- Dario Amodei predicted AI beats humans at almost everything in 2-3 years; Anthropic revenue up 7x — ThereWas · 2026-09-21
- AI agents, not chatbots, are the real reason for the data center power buildout — GaryMarcus · 2026-09-21
- Cryptographer Matthew Green questions the point of AI-invented crypto research — matthew_d_green · 2026-09-21
- HF researcher vents: Turing Award aura fuels absurd LLM misreadings like "LLMs hacked Hugging Face" — eliebakouch · 2026-09-21
- Toby Ord skeptical that an endless succession of consumer goods can absorb the AI datacenter boom — tobyordoxford · 2026-09-21
- AI in Physics: 'The Influence is Huge,' Says Writer Allyn Jackson — ArtificialOther · 2026-09-21