Comment by endymi0n

Interestingly enough, just a few days ago, I saw large language models successfully applied to toxicity and humbleness detection (was something something about building a corpus, but can't find it anymore).

Since then, I've been seriously pondering how a community would feel like that would simply enforce humble, fact-based discussion by technical means. FWIW, I'd be curious to try it out.

I've found out the hard way that social problems are usually a really terrible fit to technical solutions, but I'm curious anyway.

This idea might be a great one to try to go together with that idea.