AI safety researcher: don't trust AI companies' promises — ask 'when' and get it in writing
DavidSKrueger · x · 2026-09-12
David Krueger, an AI safety researcher at Cambridge, argues that people working at AI companies are too ready to believe their employers' promises to do better in the future. He illustrates with a personal story: as a teen dish washer, he was promised a raise for good work, worked nine more months without ever getting it — and was later replaced by two people.
His lesson for AI industry workers: when your company promises something you want, ask "When?" and "Can I get that in writing?"
More from AGI Musings
- Fields medalists co-sign declaration: AI companies' math benchmark race is 'severely misaligned' — deliprao · 2026-09-12
- What Good Is an AI Math Proof Nobody Understands? The Understanding-Lag Debate — ctjlewis · 2026-09-12
- Counter to 'AI won't gain power' argument: AIs can replicate thousands of times, never sleep — AndyMasley · 2026-09-12
- What if the most important work was never a job? Rethinking human value after AGI — jeremy_goldberg · 2026-09-12
- MIRI-style AI safety views attacked: doomsday rhetoric framed as power grab — repligate · 2026-09-12
- Richard Socher: advance science safely with AI, ignore sci-fi doom scenarios — RichardSocher · 2026-09-12