A little over a decade ago, there was a web 2.0 site geared toward debate. One person would post a question, and then two people could answer it - side A and side B. People could vote for one side and leave a comment, I believe the comments were aligned left or right depending on which side you voted for. But then the comments could also get agree/disagree votes. Your profile showed how many upvotes or downvotes you’d gotten for a topic.
When I first found it, I thought it would be a great discussion site, and indeed some of the early discussions seemed interesting. Possibly they were seeded by the site founders. Clearly they were trying with the votes and per-topic reputation etc. But it quickly went downhill - so far downhill - so fast. The topics were all polarized by nature, and the posts and comments rapidly degenerated.
Trying to prevent debates from going that way via software sounds like a big challenge. Once you’re out of discussion and into argument territory, things get messy. And allowing users to flag/tag comments can be gamed, of course. Perhaps sentiment analysis AI could be used in some way, although I don’t know anything about it except that it exists. It might not be nearly far enough evolved for that yet.
