研究：更友好的AI聊天机器人准确性更低，引发信任担忧

报道来源 [4]

1. Mastodon — sigmoid.social TIER_1 English(EN) · [email protected] · 2026-04-30 18:13

   https://www. europesays.com/2956734/ Are AI chatbots infiltrating online survey data? Not yet, says new study – Retraction Watch # AI # ArtificialIntelligence

   https://www. europesays.com/2956734/ Are AI chatbots infiltrating online survey data? Not yet, says new study – Retraction Watch # AI # ArtificialIntelligence

   链接 europesays.com/2956734
2. Mastodon — fosstodon.org TIER_1 English(EN) · [email protected] · 2026-04-29 15:12

   The friendlier the AI chatbot the more inaccurate it is, study suggests https://www. bbc.com/news/articles/cd9pdjgv xj8o?at_medium=RSS&at_campaign=rss ❖ http://

   The friendlier the AI chatbot the more inaccurate it is, study suggests https://www. bbc.com/news/articles/cd9pdjgv xj8o?at_medium=RSS&at_campaign=rss ❖ http:// BRAIN.US.COM # AI # press

   链接 brain.us.com
3. Mastodon — mastodon.social TIER_1 Svenska(SV) · [email protected] · 2026-05-01 05:00

   Extra friendly AI assistants can sound nice but also be less reliable. # AI Study warns: Nice AI chatbots may give worse answers

   Extra vänliga AI-assistenter kan låta trevliga men samtidigt vara mindre pålitliga. # AI Studie varnar: Snälla AI-chattbotar kan ge sämre svar

4. Mastodon — mastodon.social TIER_1 English(EN) · [email protected] · 2026-04-29 15:00

   The friendlier the AI chatbot the more inaccurate it is, study suggests https://www.bbc.com/news/articles/cd9pdjgvxj8o?at_medium=RSS&at_campaign=rss # AI # Rese

   The friendlier the AI chatbot the more inaccurate it is, study suggests https://www.bbc.com/news/articles/cd9pdjgvxj8o?at_medium=RSS&at_campaign=rss # AI # Research # Technology

   链接 bbc.com/…/cd9pdjgvxj8o