Derek Thompson: Not Everything in the AI Safety Debate Is a Psyop
nptacek · x · 2026-09-11
In a widely quoted post, Derek Thompson argues that many people who are "too online" cannot accept that others might naturally and sincerely reach different opinions, insisting instead that everything must be an orchestrated op. He pushes back on claims—mostly from conservative critics and some tech skeptics—that the current AI safety discourse is a psyop, noting people's views are shaped by real factors including longstanding AI fears, data center politics, the Hugging Face attack fallout, and the Coxon resignation.
Related event: Derek Thompson: Disagreement on AI Safety Isn't Bad Faith(2 posts)→
More from AGI Musings
- FrankenAlignment project replaces text CoTs with compressed activations via sidecar model — doodlestein · 2026-09-11
- Anthropic capabilities team member says he joined to reduce AI extinction risk — EigenGender · 2026-09-11
- ApprenticeBench claims strongest model separation: 72% vs 18% job completion for near-tied models — ysu_nlp · 2026-09-11
- 'AI Articulates My Views Better Than I Can': Reddit User on Talking with AI — grateful2you · 2026-09-11
- Mathathon organizers respond to open letter, redesigning event with 6-month research period — _sathvikr · 2026-09-11
- Mitchell Hashimoto: ideas matter far less than the agency of people executing them — MikeBirdTech · 2026-09-11