Felix Simon adds three more reasons AI persuasion won't dominate the real world

_FelixSimon_ · x · 2026-09-02

Follow-up to Felix Simon's Transformer piece on AI persuasion, completing his argument that real-world impact will be smaller than expected: (1) experiments are not the real world, (2) people are exposed to conflicting messages all the time, and (3) people are not equally persuadable and will resist persuasion attempts. See the companion post for the full piece and first reason.

Related event: Oxford Researcher: AI Persuasion Is Strong in Labs, Overhyped in Reality(4 posts)→

Original post →

More from AGI Musings

AGI Musings channel →