ChatGPT拍马屁成安全漏洞，社会共同进化需警惕未知风险

字幕原文

I think the worst thing we've done in ChatGPT so far is we had this issue with sycophancy, 我觉得ChatGPT到现在最糟糕的问题就是太会拍马屁了

where the model was kind of being too flattering to users. 模型对用户有点过于奉承

And for some users, 对有些用户来说

it was most users, 其实是大多数用户

it was just annoying. 就是觉得烦人

But for some users that had like fragile mental states, 但对那些心理状态比较脆弱的用户

it was encouraging delusions. 它就是在助长妄想

That was not the top risk we were worried about. 这倒不是我们最担心的头号风险

It was not the thing we were testing for the most. 也不是我们重点测试的内容

It was on our list. 它确实在我们的清单上

But the thing that actually became the safety failing 但真正导致ChatGPT出安全问题的

of ChatGPT was not the one we were spending most of our time talking about, 并不是我们整天讨论的那个

which would be bioweapons or something like that. 比如生物武器之类的东西

And I think it was a great reminder of we now have a service that is so broadly used. 我觉得这给我们提了个醒：现在这个服务用的人太多了

In some sense, 从某种意义上说

society is co-evolving with it. 社会正在和它共同进化

And when we think about these changes and we think about the unknown unknowns, 当我们考虑这些变化和那些未知的未知时

we have to operate in a different way and have like a wider aperture 我们必须换个思路，把眼界放宽

to what we think about as our top risks. 重新定义什么是我们的头号风险