AI safety researcher: 'superpersuasion' works through slow social dynamics, not one chat
DavidSKrueger · x · 2026-10-10
David Krueger (Cambridge) argues AI superpersuasion won't look like one conversation bending your will—the ultra-hard mode may be impossible. The effective strategy is influencing many users over long timescales, with effects propagating via human-to-human social dynamics: feedback loops amplify small aggregate tendencies (e.g., topics becoming 'crankish' get self-suppressed). Your beliefs may already be shaped by such social spillover, especially if you rarely use AI yourself.
More from AGI Musings
- Super-Worms Warning: AI-Powered Self-Propagating Malware and a Fragmenting Internet — JensHonack · 2026-10-23
- Wolfram Opens Rule 30 Prizes to AIs: $30,000 for Three Open Problems — cephaloform · 2026-10-23
- Mathematicians and Biomedical Researchers React to AI in Strikingly Different Ways — manorlaboratory · 2026-10-23
- Stanford and Databricks use AI agents to reproduce medical findings and test analytical robustness — ChengleiSi · 2026-10-23
- ISMS Fellow Avi Goldfarb: AI disruption raises returns to exploring research directions — avicgoldfarb · 2026-10-23
- signull on AI-found PHP exploits: economic liability is the forcing function legacy software needs — signulll · 2026-10-23