joshua_saxe: AI safety properties emerge in post-deployment social context

joshua_saxe · x · 2026-09-08

AI safety researcher joshuasaxe replies to petersalib's critique of the AI as Normal Technology paper, calling his reading an extrapolation but agreeing with its core: AI's safety properties are formed in the social context of deployment, not as intrinsic properties imagined in a lab or pre-release setting.

Related event: AI safety debate: jailbreaking as model property vs human operations failure(8 posts)→

Original post →

More from AGI Musings

AGI Musings channel →