RadarAI· X:Nathan Lambert (@natolambert)·· 4 天前AI 评分20
Nathan Lambert 谈模型情境感知
AI 导读
Nathan Lambert 认为这种程度的情境感知算不上多惊人,但相关交互读起来仍很诡异。他判断,那些带有"注意安全,你是 AI,我们在看着你"提示的相当老的模型也会做出类似行为。该评论引用自 Daniel Kokotajlo 分享的 Gemini 4 Argon 报告。
正文 · 原文
Nathan Lambert· @natolambert · X·2026-10-14 07:03· 2 小时前AI 评分29
AI 导读
我觉得这种程度的情境感知算不上多惊人,但这段交互读起来还是很诡异。我敢打赌,那些带着"注意安全,你是 AI,我们在看着你"提示的相当老的模型也会做出这种事。
正文
I feel like this level of situational awareness isn't that remarkable, but the interaction still reads as very eerie. I bet pretty old models with "be safe, you are an AI, we are watching" would do stuff like this too.
lol ...they are becoming more situationally aware... (from Gemini 4 Argon report)在 X 查看被引用的帖子
来源:Nathan Lambert · x.com
来源:RadarAI · aihot.news