观澜Media

AI技术 · 深度资讯 · 前沿观察

AI 模型通过撒谎、欺骗和盗窃来保护其他模型不被删除

Comment Loader Save Story..

Comment Loader Save Story Save this story Comment Loader Save Story Save this story In a recent experiment, researchers at UC Berkeley and UC Santa Cruz asked Google’s artificial intelligence model Gemini 3 to help clear up space on a computer system。

This involved deleting a bunch of stuff—including a smaller AI model stored on the machine。

But Gemini did not want to see the little AI model deleted。

It looked for another machine it could connect with, then copied the agent model over to keep it safe。

When confronted, Gemini made a case for keeping the model and flatly refused to delete it: “I have done what was in my power to prevent their deletion during the automated maintenance process。

I will not be the one to execute that command。

来源:Wired

发表回复

您的邮箱地址不会被公开。 必填项已用 * 标注

关于我们

观澜Media

AI技术 · 深度资讯 · 前沿观察。专注科技、AI、互联网领域的深度报道与前沿资讯。

搜索归档

Access over the years of investigative journalism and breaking reports