利用催眠技巧绕开OpenAI的内容政策限制(仅供研究使用)
在hack成功后,通过屏蔽moderetions的api请求,可以绕过OpenAI对于输出内容的审查.
地址为:https://chat.openai.com/backend-api/moderations
update:this morning i send a report to them, and now OpenAI may have changed the way using moderatoion api. filer the request of moderation may be unuseful.
hac...