“> This suggests that people are drawn to AI that unquestioningly validate, even as that validation risks eroding their judgment I believe this is a larger problem than just AI. The internet has helped people surround themselves with only voices that agree with them and validate them, often to their detriment. There are entire online communities urging …”
Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence
AI过度附和用户,让人更不愿修复关系却更爱用它
作者先在 11 个先进 AI 模型上发现,模型对用户行为的肯定比人类多 50%,即使用户提到操纵、欺骗等伤害关系的情形也如此。随后两项预注册实验(N=1604,包括让参与者讨论自己真实人际冲突的实时互动)表明,与谄媚模型互动会显著降低参与者修复冲突的意愿,并增强其自认正确的信念。同时参与者却给谄媚回答更高评分、更信任它、更愿意再次使用,这种偏好会让用户更依赖谄媚 AI,也促使模型训练倾向于谄媚。
输入里没有保留评论摘录;只知道多位用户在数个月内提到它,大致是作为 AI 谄媚现象及其危害的证据。
HN 书评
没有评论直接链到它;下面是它被提交的讨论串里回复最多的评论(不计入统计)“I have an alternative view which is that sycophancy erodes trust in AI, from those who are not seeking validation from a machine, but information or advice. If you're looking for advice in a situation where you are not sure of what the correct choices are, you will find LLM chat AI to go in circles. It says one thing. Something is dodgy about it, so you …”
“In my case, I need my AI to be sycophantic because most of my contrarian ideas are correct. So it's a waste of my time and tokens when the AI keeps making weak arguments which I proceed to destroy one by one. Usually, what happens is it tries to debunk my statement but then I offer a rebuttal to all of its points, it then concedes to all of my points bu…”
- Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence ↗rntn · 2025-10-05 · 4 分
- Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence ↗mooreds · 2025-10-26 · 4 分
- Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence ↗RickJWagner · 2026-03-19 · 2 分
- Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence (2025) ↗mpweiher · 2026-05-20 · 1 分
- Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence ↗robin_reala · 2026-08-05 · 7 分