AI safety researcher: AI has always gained powerful capabilities because humans freely hand them over

DavidSKrueger · x · 2026-09-20

David Krueger endorses an observation about AI safety: throughout history, whenever AI has gained access to some new powerful capability, the answer has almost always been "someone freely gave it to them" — a pointed critique of frontier release and open-source culture rather than any AI takeover scenario.

Original post →

More from AGI Musings

AGI Musings channel →