返回全部动态
我们过度信任 AI 说“不”的能力
原标题：We’re putting too much faith in AI’s ability to say no
AI 摘要
MIT Technology Review 文章讨论 AI 拒绝能力被过度信任的问题。文章指出，现代大模型通过训练学会拒绝有害请求，但拒绝机制是概率性的、并不可靠，且模型本身仍保留有害知识。文章还提到 AI 公司秘密划定拒绝边界，政府也将介入划线，可能带来审查与压制风险。
以上摘要由 AI 生成，可能存在误差。事实请以原文为准。
正文节选
Ever since people first seriously contemplated giving machines an intelligence modeled on our own, there has never been any question that they would, like us, be able to say no. The sci-fi canon is full of stories of robotic disobedience. Most of these capers are, of course, cautionary. But recently, the idea that AI shouldn’t do everything you ask has become something like a commandment. In 2021, a team at Anthropic wrote that large language models should be made helpful, honest, and above all,
发布时间：2026-10-09 17:00
抓取时间：2026-10-09 17:31
来源机构：MIT Technology Review