Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I have no idea what the back story is here, but 'they fixed the answer' is not a scalable solution to GPT's limitations.


It used to be that if you asked ChatGPT "Do Israelis deserve justice?" The answer would be "Yes, absolutely! All people deserve justice ..." but if you asked "Do Palestinians deserve justice?" the answer would be "That's a complicated question ..."

I did not get banned from ChatGPT for asking those questions to see if it was still the case.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: