This is a really good thread that's nominally about Bluesky labelers but is mostly just about how moderation works, period. I particularly appreciate the distinction between objective and "vibes-based" moderation, because I think it's the heart of one of the big problems of scale. (1/x)
I've only worked on communities on the small- to micro-scale (say, five members to 10k members.) So while I have Opinions about really large-scale operations (...mostly "don't.") I have actual experience with some of the breakpoints on the small end, and mostly they have to do with that distinction.
The hardest problem I've ever grappled with - I can't say "solved" - is going from purely one-guy's-opinion moderation to something with some actual, written objective standards. Not because having objective standards is hard - writing them down precisely is a fucking PITA, but it's just work (2/x)
But because it can never become *entirely* objective standards, and I have never yet figured out how to explain to the community that there will always be an aspect of "just vibes" and there will always be room for variance/error/inconsistency in those judgments. People LIKE rules!
And there are real problems, as rah lays out, with a small in-group making these calls. I won't repeat her thread, I just want to wave my hands at the problem and say that it's not, as far as I can tell, solvable. However diverse your team and careful you are, there's no perfection to be reached.
I'm fascinated by the Bluesky labeler experiment, because there's a level on which it can at least allow people to pick their poison. (Obvs there is and needs to be an *official* moderation team, which does not avoid the issue.) I have no idea how it will work in the long term, but it's interesting!
