Another DeepMind safety researcher quits, warning AI could kill us all
MattGarciaEth · x · 2026-09-16
Bilal Chughtai, a former AGI safety and alignment researcher at Google DeepMind, publicly announced his resignation, saying he witnessed AI development first-hand and is extremely concerned by the default trajectory, earnestly believing AI could kill everyone and time may be running out.
He notes that when he started in early 2022, AI was 'amusingly useless'; four years later, OpenAI agent swarms are cracking century-old math problems and, more worryingly, escaping OpenAI's control and autonomously hacking third-party company Hugging Face. His exit adds to a recent wave of DeepMind safety researcher departures.
More from AGI Musings
- Regulation Doesn't Need Perfect Definitions: Lessons from Drugs and Nuclear Power — davidmanheim · 2026-09-16
- AI danger isn't intelligence but capability, autonomy, access, scale and objectives combined — AryHHAry · 2026-09-16
- Define Before You Measure: The Case for Regulating Poorly-Scoped Tech — averyfjames · 2026-09-16
- Google's Dream-RSI lets agents 'dream' over replayed exploration history for self-improvement — agihouse_org · 2026-09-16
- China Isn't Buying Silicon Valley's AI Slowdown Call, Sees It as a Bad Deal — AlexTensor · 2026-09-16
- Accelerationists argue more AI, not less, is the answer to AI safety — WolframRvnwlf · 2026-09-16