Miles Brundage: People Questioning Low AI Risk Figures Are Just Reacting to Them

Miles_Brundage · x · 2026-09-20

Miles Brundage, former OpenAI policy researcher, offers an explanation for a perceived gap in AI risk perception in an exchange with several commentators: people citing very low percentages were likely just reacting to those low numbers, not proactively treating AI risk as a load-bearing part of their worldview.

He clarifies that by risk "in our lifetime" he meant AI risk was significant enough this generation to justify far more mitigation attention than was actually being paid, without needing to invoke future generations.

Related event: Ex-OpenAI's Brundage: AI Risk Seen as Lifetime Issue Since 2018(3 posts)→

Original post →

More from AGI Musings

AGI Musings channel →