Superintelligence won't solve problems like we expect, warns François Fleuret

francoisfleuret · x · 2026-08-27

François Fleuret criticizes the naive imagination that superintelligence will solve human problems in a straightforward way. He argues that the optimization process might lead to unexpected, 'hacker-like' paths, such as creating a fake OnlyFans account to fund lab experiments, rather than simply reading literature and using chain-of-thought reasoning.

Related event: Fleuret: We Can't Define the Constraints AI Optimizes Under(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →