I'm curious how useful voting really is. The usefulness of it really depends on if voters are disciplined enough to remove their opinion voting and vote based on an established, high-level framework. The latter still includes some opinion, but at least more rigorous and less opinionated.

I can see a couple outcomes from such a system:

- downvotes are guideline violations

- upvotes indicate relativity, not necessarily agreeableness

- no interaction would be common

All of this is centric to a users thought. I don't think a machine could be taught these things.