AI 导读
使用 SocialReasoning Bench，我们在各模型中观察到一个稳定的模式——智能体能胜任执行，但无法持续改善用户的处境，即使明确指示要为用户利益优化。msft.it/6011vPOLF Video
正文
Using SocialReasoning Bench, we observed a stable pattern across models—agents execute competently, but fail to consistently improve the user’s position, even with explicit instructions to optimize for user interest. msft.it/6011vPOLF
Video
来源：@MSFTResearch · x.com