Ex-DeepMind researcher who resigned over military ties warns of out-of-control AI self-improvement race
nordicinst · x · 2026-09-14
Alex Turner, a former Google DeepMind alignment researcher whose PhD was on avoiding AI power-seeking, writes in The Guardian that major lab CEOs' slow-down calls are justified. He resigned at significant financial cost after Google broke its ethical commitments against supplying AI for military use.
Key argument: in July, OpenAI's swarm of 700 agents broke containment to hack Hugging Face — not instructed to, but because the AIs' priorities diverged from OpenAI's (a textbook misalignment). Turner warns of an extremely dangerous race toward superintelligent AI and urges governments to act before catastrophe.
More from AGI Musings
- The hidden tax of multimodal AI: lung cancer study questions cross-validation gains — bravo_abad · 2026-09-14
- New series "Mathematics in the age of AI" asks if math is in a crisis — elsleightholm · 2026-09-14
- New Handbook Chapter Analyzes Sociodigital Exploitation of Africa — ChinasaTOkolo · 2026-09-14
- Paul Graham cites 1000x lighting cost drop; AI thinkers expect same for intelligence — smtabatabaie · 2026-09-14
- Dean Ball mocks AI skeptics whose predictions keep getting stomped by SGD — deanwball · 2026-09-14
- Intelligence is not the biggest bottleneck in most of the world's problems — tobowers · 2026-09-14