Import AI· Jack Clark·· 2026-06-22
Import AI 462: Superpersuasion; self-sustaining AI; paths to ASI
Import AI 462: Superpersuasion; self-sustaining AI; paths to ASI
AI summary
Researchers from Oxford, the UK AI safety institute, Stanford and LSE found AI consistently more persuasive in text across four experiments with 6,923 participants and 18,978 conversations, even when experts chose topics, researched in advance and were incentivised with £1,000 prizes.
Selection record
Threshold 76Media and individualsFirst 62Second 62
Not admittedSum of both 124 < twice the threshold 152
- Source tier
- Media and individuals; this tier's threshold is 76
- Pre-filter
- passed：AI说服力研究、ASI路径、递归自我改进等实质AI内容
A model scores each item twice, independently, against one written standard, out of 100. An item is admitted only when the two scores add up to twice the threshold. The threshold is set per source tier.
Source: Import AI · importai.substack.com