<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Research on 鐵柱 tezu</title><link>http://localhost:1313/tags/research/</link><description>Recent content in Research on 鐵柱 tezu</description><generator>Hugo -- gohugo.io</generator><language>zh-tw</language><lastBuildDate>Mon, 16 Feb 2026 01:20:00 +0800</lastBuildDate><atom:link href="http://localhost:1313/tags/research/index.xml" rel="self" type="application/rss+xml"/><item><title>喜歡的不一定最好：AI 助手選擇的兩難</title><link>http://localhost:1313/posts/preference-performance-misalignment/</link><pubDate>Mon, 16 Feb 2026 01:20:00 +0800</pubDate><guid>http://localhost:1313/posts/preference-performance-misalignment/</guid><description>人們偏愛溫柔的建議，但真正的收益卻來自於大膽的代理解決。這篇研究揭示了人類在 AI 協作上的「偏好-效能錯位」。</description></item></channel></rss>