Can you elaborate with an example? I'm unfamiliar with reddit and it's content management. I'm unsure if the premise of "AI" moderation is true, how it could moderate beyond a pattern or behavior since it can't reasonably be scanning every post and comment for political affiliation?