News for hackers. Stuff that matters.

A comment on Building a no-code toxicity classifier by talking to GitHub Copilot. Parent comment

Re:Building a no-code toxicity classifier by talking to GitH... by Spivak on Friday March 25, 2022 @03:21PM (#30806117)
Would you say the same about a less divisive speech pattern like a flamebait/flamewar classifier? Because it’s already double with simple heuristics like upvote/comment ratios and seems like a fine fit for a moderator assisted classifier.
[ Parent | on HN ]
Re:Building a no-code toxicity classifier by talking to GitH... by IshKebab on Friday March 25, 2022 @04:12PM (#30806586)
Yeah I would say the same. It's also not well defined or black or white but that doesn't mean you have to just give up. You can do better than nothing in both cases.
[ Parent | on HN ]