Researchers who turned down Anthropic and OpenAI: choosing moral ambition over lab wealth
AaronBergman18 · x · 2026-10-29
A threaded reflection on career choices in AI safety: researcher JacquesThibs explains he quit mechanistic interpretability in early 2023 because he expected Anthropic and other labs to invest heavily in interp anyway, and his safety startup struggled not from lack of funding but from doubt that the direction would create high moral value. Another researcher notes that joining an AI lab would have made building generational wealth easy, but he'd rather instill moral ambition, courage and integrity in his future children.
More from AGI Musings
- A satire: AI comes for math one field at a time, until it reaches elliptic curves — matthew_d_green · 2026-10-29
- Musk endorses Larry Page's prediction: Von Neumann probes exploring the galaxy by 2100 — elonmusk · 2026-10-29
- Princeton's Tom Griffiths at MAIN 2026: mapping AI's jagged edges with cognitive science — karimjerbineuro · 2026-10-29
- Lawyer predicts AI-displaced big law talent will pivot to smart contracts and in-house negotiation — curious_vii · 2026-10-29
- Greg Kamradt: Last 12 months were about games, next 12 about open-ended innovation — GregKamradt · 2026-10-29
- Reddit compares Musk's 'Universal High Income' to a prisoner's flying horse parable — Kitchen-Trouble7588 · 2026-10-29