Essay: The mundane AI risk is humans happily ceding control as institutions hollow out

anshulkundaje · x · 2026-09-19

Ruxandra Teslo's new essay argues the most realistic AI danger isn't existential risk but voluntary abdication: as human cognitive capacities atrophy and the institutions meant to cultivate them hollow out, we may actively prefer handing control to machines.

She cites the recent letter from mathematicians worried about AI being dismissed as "protectionist" as a canary in the coalmine, and critiques optimistic narratives that assume humans will retain goals and taste while machines do cognitive work.

Original post →

More from AGI Musings

AGI Musings channel →