Ex-DeepMind researcher testifies Google, including Demis Hassabis, backed off AI safety commitments
BlackHC · x · 2026-10-06
Former Google DeepMind researcher Alex Turner gave powerful testimony at a NYC hearing on AI, describing how his push for Google to commit to AI safety failed — with Demis Hassabis included, the company ultimately walked back earlier commitments. BlackHC adds that this shows why voluntary self-governance of big corporations shouldn't be trusted, and how impossible it proved to keep Google from "being evil."
More from Companies & People
- Debate swirls over OpenAI's unreleased math proofs as attitudes flip in two weeks — basedjensen · 2026-10-06
- NotebookLM lands at No. 9 in a16z consumer AI rankings as Google takes 4 of top 20 — stevenbjohnson · 2026-10-06
- O(n²) matrix multiplication is almost certainly false even if ω = 2, says basedjensen — basedjensen · 2026-10-06
- Kevin Mandia: AI attacks run at machine speed, defense must go autonomous — a16z Podcast · 2026-10-06
- Jobs Called Touchscreen Macs 'Terrible' — 16 Years Later, Apple May Ship One — Kiteaton · 2026-10-06
- Redditor Publishes Free, Independently Researched Deep-Dive Breakdown of OpenAI — Roadtochessmaster · 2026-10-06