With content moderation being such a controversial topic, I think AI probably presents a really unique opportunity to create an objective arbitrator to act as a filter between users and platforms (rather than a bot with hard coded triggers and lack of nuance). Maybe instead of agreeing on every outcome, we instead agree on the underlying training data and methodology implemented to create an AI which passes down a preliminary decision and outlines why the decision was taken (which could then be appealed). Anyone could suggest improvements, and prove them through simulated results.

One downside is that if it's public, anyone can figure out how to work around it to post inappropriate content to the platform before a solution is worked on, but you could have specific ways of dealing with content that is getting flagged by a lot of users (get human perspective or run through other AI designed for this).

Show more