AI safety researchers lament self-regulation trap: too soon in 2023, too late in 2026
tegmark · x · 2026-11-17
David Krueger, amplified by Max Tegmark, mocks the AI safety community's bind: in 2023 it was 'too soon to stop AI, so we settled for self-regulation'; by 2026 it's 'too late to stop AI, so we settle for self-regulation again.'
Quoted context from Kelsey Tuoc: as Bay Area practitioners increasingly expect recursive self-improvement to arrive sooner, many have given up on legislation or bilateral treaties, believing regulation simply can't move fast enough.
More from AGI Musings
- Debate: one big model is offline RL, specialized models are online RL — willcb · 2026-11-17
- Domingos on fast takeoff: 'The world is a jungle, not a racetrack' — pmddomingos · 2026-11-17
- willcb: RLaaS is just a stepping stone toward continual learning, the real AGI bet — willcb · 2026-11-17
- Domingos: recursive self-improvement has existed since the 1950s — pmddomingos · 2026-11-17
- Black Friday will be the moment consumer agents go mainstream and botting becomes everyone's game — omooretweets · 2026-11-17
- AI founder tengyanAI: the panic on X comes from people who never lived through cycles — tengyanAI · 2026-11-17