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

Is anyone working on ML to determine possible avenues of response and debate given a prompt?


I would love to see some social networks silently implementing something that rewrites comments in a more diplomatic manner when displayed to everyone else.


That would be an interesting alternative to shadowbanning, especially if it came with a label ("The comment has been autotuned.") I'd like to see them offer the submitter some alternatives or provide some analysis before they post ("Please avoid sealioning").




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

Search: