AI safety researcher flags contradiction in skepticism about pausing AI
DavidSKrueger · x · 2026-11-09
AI safety researcher David Krueger highlights a tension in the AI safety community: many people who are otherwise progress/Pinker-pilled are skeptical that pausing AI helps, because they doubt we'd use longer timelines to reduce P(doom). He finds this strange — believing humans generally use time well implies longer timelines should also help reduce risk. A substantive contribution to the pause-AI strategy debate.
Related event: Cambridge Scholar Questions AI Safety Community's Stance on Pausing AI(2 posts)→
More from AGI Musings
- OpenAI engineer: abundant intelligence will bring safer cars, cheaper homes — OfficialLoganK · 2026-11-09
- How to win the AI narrative: cures, cheaper living, and data centers people can accept — nbaschez · 2026-11-09
- Patrick Collison on the hard question of AI's positive scenario: work, scarcity, equilibrium — mhmazur · 2026-11-09
- Pedro Domingos: every journal should train AI on its full submission corpus as first-stage reviewer — pmddomingos · 2026-11-09
- Reminder: AI is a perpetual bullshit machine that turns us against each other — gnukeith · 2026-11-09
- A perfect human: notes on making every LLM user smarter with each use — keyanzhang · 2026-11-09