Demis Hassabis proposes a standards body to test frontier models and coordinate slowdowns
TheMoonMidas · x · 2026-09-14
The "conditional approach" in the AI slowdown debate comes from Demis Hassabis, whose July essay A Framework for Frontier AI proposes:
- AGI is likely only a few years away, with impact perhaps 10x the Industrial Revolution at 10x the speed
- A standards body should test frontier models and could coordinate a slowdown if risks justify one
- Smaller, non-frontier models would be exempt
The post also cites OpenAI chief scientist Jakub Pachocki: no lab has solved alignment and monitoring well enough to keep scaling at maximum speed much longer; he favors voluntary slowdowns and shared safety standards.
More from AGI Musings
- Yoav Goldberg translates AI industry speak: 'third-party evaluators' are spies, 'pacing' means stop spending — yoavgo · 2026-09-14
- zetalyrae: true atheists are rare — people find teleology everywhere, even thermodynamics — zetalyrae · 2026-09-14
- Humans got better at chess and Go after AI dominance — and math may see the same effect — RexDouglass · 2026-09-14
- Pundits Screamed 'Communism' Over a Short Letter They Barely Read — menhguin · 2026-09-14
- Melanie Mitchell: 'rogue AI swarm' headlines are misleading metaphors amplifying real risks — anilkseth · 2026-09-14
- After the OpenAI hack: Bengio on cheating agents, Sacks fires back at Dario's 'pace the frontier' — RemiCadene · 2026-09-14