High p(doom) at Anthropic and OpenAI is baked into their founding culture, not a random sample
robleclerc · x · 2026-09-11
The author argues researchers at Anthropic and OpenAI didn't arrive at high p(doom) recently — the fear is embedded in their founding DNA: OpenAI was created out of fear Google would get to AGI first, and Anthropic split off over safety concerns. That's why these researchers, and Chinese labs, don't uniformly hyperventilate about AI risk.
Related event: High p(doom) Is Baked Into OpenAI and Anthropic's Culture(3 posts)→
More from AGI Musings
- Anthropic researcher defends 50% doom estimate as critics call it fraud — TinfoilTricorn · 2026-09-11
- Economist Ben Moll: under 0.01% odds airlines vanish to AI robots by 2030 — soumitrashukla9 · 2026-09-11
- Ex-OpenAI formal methods lead ships Code Contracts: agentic, non-formal code verification — TacoCohen · 2026-09-11
- Andrew Ng Names the Profession AI Hits Hardest — Openings Are Still Rising — cen6wkf · 2026-09-11
- Mathematicians Confront the AI Apocalypse — scientificamerican · 2026-09-11
- Mathematicians Confront the AI Apocalypse — scientificamerican · 2026-09-11