AI safety row: did Scott Alexander outline Anthropic's plan to release an aligned superintelligence?
ohabryka · x · 2026-11-22
ohabryka quotes a Scott Alexander post (with screenshot) arguing that "Plan C" — where alignment-pilled AI labs coordinate a mutual pause, burn their lead on alignment research, and release a hopefully-aligned superintelligence at the last possible moment if a global pause fails — reads like a plan for world takeover with an aligned superintelligence. He says Anthropic employees praised the post, one refusing to deny it on confidentiality grounds; catherineols disputes this, saying Anthropic staff liked the critique of questionable assumptions, not the takeover thesis.
Related event: Anthropic Accused of Plotting Aligned Superintelligence Takeover(2 posts)→
More from AGI Musings
- Anthropic researcher: build state AI capacity instead of arguing about doom — joshua_saxe · 2026-11-22
- You can only guarantee zero mistakes by making zero predictions — burny_tech · 2026-11-22
- The 'Agent Platform of 2027' blueprint: actors, RBAC filesystems, inbox UX — yawnxyz · 2026-11-22
- AI practitioner: I can no longer predict AI capabilities with confidence — burny_tech · 2026-11-22
- 'There's never been a better time to go into math' amid AI boom — burny_tech · 2026-11-22
- When an AI agent claims it solved something, what evidence would actually convince you? — Zealousideal-Pop4857 · 2026-11-22