AI is a worryingly-good persuader—but real-world influence is harder, argues Oxford researcher
_FelixSimon_ · x · 2026-09-02
Oxford researcher Felix M. Simon writes in Transformer that while the latest AI models are remarkably persuasive—recent preprints show chatbots can out-argue seasoned campaigners at getting people to part with money, at least in experiments—turning that ability into meaningful real-world influence is much harder. He distinguishes persuasion (giving inspectable, rejectable reasons) from manipulation, notes the two are easily conflated, and argues that fears of AI nudging masses are premature even as platforms and governments eye the technology for counter-conspiracy efforts.
More from AGI Musings
- AI commoditization: founders should seek friction, ops and regulatory moats — vaibhavbetter · 2026-10-18
- Anthropic's Jack Clark: from 10M scientists to 10 billion virtual ones — coherence · 2026-10-18
- Paras Chopra: Intelligence picks good options, power defines the option space — paraschopra · 2026-10-18
- Terence Tao on 'War Footing' as OpenAI Claims Navier-Stokes Solve Shakes Mathematics — AlexTensor · 2026-10-18
- Anthropic report: junior technical hires fell from 27% to 9% as agent hours hit 6x human hours — rohanpaul_ai · 2026-10-18
- Redditor proposes AI regulation framework: text detection, face verification, open algorithms — techtotechbytechy · 2026-10-18