Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Study: LLMs are prone to social identity biases similar to the way humans are (nyu.edu)
1 point by giuliomagnifico 8 months ago | hide | past | favorite | 1 comment


But fortunately:

> AI biases can be reduced by carefully selecting the data used to train these systems




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

Search: