DeepMind's Irving calls to halt AI progress now, with treaties to follow
davidmanheim · x · 2026-10-23
DeepMind safety researcher Geoffrey Irving argues AI progress should stop now — first unilaterally by key AI companies, then across more US companies, and finally via international agreements including China. The reason isn't that current models could kill everyone, but that coordination takes time, and acting immediately buys time for later steps.
His article concludes that the necessary moratorium on dangerous uses and model development, plus accompanying governance regimes, will take time to negotiate — so concrete steps must start today even as extreme risks grow.
Safety researcher David Manheim applauded the statement but called it still too optimistic: single governments move slowly and treaties are slower, which is why he's been urging treaty negotiations to start well before risks appear, for at least three years.
Related event: DeepMind Safety Lead Calls for Halting AI Progress Now(2 posts)→
More from AGI Musings
- "OpenAI/Anthropic will ditch users and solve quantum gravity" take gets mocked — QuintinPope5 · 2026-10-23
- Banks may ban AI agents on security grounds to fend off agentic bank runs — Gccooke · 2026-10-23
- People who explain complex tech clearly may become more valuable than those who build it — signulll · 2026-10-23
- Dev warns labs must fix agent-security gap before thousands-of-agents attacks — imjustnewatai · 2026-10-23
- Daniel Faggella's posthuman take: abandon the human condition or lose causal purchase on the future — DKokotajlo · 2026-10-23
- David Patterson predicts superintelligence will hit IQ 160 by end of next year — davidpattersonx · 2026-10-23