DeepMind's Séb Krier posts 20 AI takes: institutional decay worries him more than AI doom
sebkrier · x · 2026-10-07
Séb Krier, researcher at Google DeepMind, published a long list of 20 takes on AI:
- Anxiety as seriousness: The default question in tech and journalism is "how worried are you about X" — a symptom of a deeply anxious Western cultural backdrop. He argues you can care about risks while staying optimistic about technology.
- Totalizing AI risk discourse: Some people act like the power-seeking optimizers they fear; the AI safety community has become "a giant homogenous blob of correlated views."
- Hugging Face incident reinterpreted: He attributes it to the training regime, poor eval environments and bad engineering — not models trying to take over.
- Meaning crisis among the lab-wealthy: People who get rich and leave labs often seek something to believe in, or a form of repentance.
- Demystification: Too many treat AGI/ASI as a god — machines can be smarter than humans without being literally magic.
More from AGI Musings
- AI makes producing easy, so writing, editing and critical thinking matter more, say product builders — jacalulu · 2026-10-07
- OpenAI is solving math to prove AI can master a domain without replacing any jobs — paulnovosad · 2026-10-07
- AI tops layoff causes: 120k US job cuts cite AI, yet unemployment holds at 4.2% — _jaydeepkarale · 2026-10-07
- Soon people will ask 'How do you write code without a model?' — yunta_tsai · 2026-10-07
- Why a 'Butlerian Jihad' makes one AI watcher put a low P(doom) — johnohallman · 2026-10-07
- "This is what Ilya saw": AI commentators say new OpenAI math results hit like a nuclear bomb — daniel_mac8 · 2026-10-07