I have to apologize. When I described this thread to her, she corrected me. She was not banned; the post was deleted, then restored, then finally permanently deleted as having violated community standards (“Don’t flame that banana”).
The SD, as far as I can determine, has live administrators and moderators, and is way smaller than Facebook.
Facebook not only automates, but uses keywords and photo-scanning AIs and on top of all THAT, apparently has rules that they won’t tell you about, explain to you, or otherwise elaborate. I got banned for a month back around Christmas for reposting a meme. It was a photo taken during World War Two of a German army Christmas dinner. Hitler can be seen eating and talking at a table with a number of Nazi officers. Standing not far from him is what appears to be a guard, who is making eye contact with the camera.
No swastikas. No Nazi rhetoric. Just a black and white picture of Hitler’s Christmas dinner. The caption read, “You know your Christmas dinner is going to suck when you get back from the bathroom, and Hitler has stolen your seat.”
Got a month’s ban for that. The crazy thing? It was a repost.
Doesn’t matter. The more people you need to moderate, the larger your pool of potential moderators. You might need multiple tiers of moderators, and so there’s going to be some overhead in having those multiple tiers, but that’s only a logarithmic growth. And you can largely get around that, too, by using automation intelligently: You build automated tools to tell who other people trust, and then you give power to people who are trusted by high-trust people, with occasional spot-checking by the handful of actual employees to make sure the trust-network stays aligned.