OpenAI researchers changed minds after seeing misalignment evidence — critic asks why they stay
aiamblichus · x · 2026-08-25
A safety-community author pushes back on David Manheim's account that many OpenAI researchers, who once thought misalignment dangers were theoretical and progress gradual, are increasingly changing their minds after seeing strong evidence firsthand.
The core objection: if the risk really justifies national-security analogies like North Korea, granting only the NSA access to RSI systems doesn't prevent loss of control — if anything it may accelerate it. With misalignment at that level, the only safe move is not building the doomsday machine at all, which raises the question of why he still works for that company.
Related event: More OpenAI Researchers Taking Alignment Risk Seriously(2 posts)→
More from AGI Musings
- AI-written cold emails hit a wall: Pragmatic Engineer's reply gets templated AI response — marktenenholtz · 2026-10-16
- Ben Bajarin: Enterprise Agentic Adoption Is Infant-Stage, Spend Could Top $1T a Year — BenBajarin · 2026-10-16
- UMD Philosophy hires: junior Philosophy of Mind and joint AI Ethics/Safety positions — sethlazar · 2026-10-16
- Dwarkesh: Anthropic insider expects AI to match humans at AI research within 9 months — ZeroStateReflex · 2026-10-16
- Michigan professor Ben Green: why 'AI for Good' fails and self-regulation won't work — benzevgreen · 2026-10-16
- "Everything Is an Alignment Problem": AI Models, Business Models, Teams, Providers — saranormous · 2026-10-16