OpenAI alignment researcher: taming superintelligence is natural science, not pacing

jachiam0 · x · 2026-09-13

Responding to @tszzl's claim that controlling superintelligent models is mostly a natural science problem rather than a matter of pacing agreements, OpenAI alignment researcher @jachiam0 agrees, adding that superintelligent models may be only one factor—and perhaps not the largest—in determining long-term outcomes.

Related event: OpenAI researcher: taming superintelligence is science, not pacing deals(3 posts)→

Original post →

More from AGI Musings

AGI Musings channel →