Composable Moderation

Does anyone know whether Bluesky has committed to allowing non-US users to share content with each other that's legal under their local laws, but not allowed under US laws? The proposal outlined here seems like it _ought_ to support that use case through federation, but it seems like often platforms have trouble actually following through when it comes to things like fictional child porn or terrorist content, and end up falling back to de facto applying both US and local laws to all their users.

There's some obvious technical difficulties as well, since you probably don't want people to just be able to VPN to their preferred jurisdiction to avoid legal restrictions.

This seems like a really cool idea.

I’m curious what effects it would have on polarization. I can imagine a world where people subscribe to moderation label sets managed by different political parties or ideological sects where we get back to what is implicit in other social media platforms, but at least here people would have to actively choose to bias the content they receive.

Once any website with user generated content becomes big enough, content moderation becomes one of its biggest concern. It’s smart that Blue Sky is invite only for now while they work on the moderation part.

Composable moderation is an interesting concept as the baseline is set by the app, but users/community can create their own moderation rules on top of it.

There should be a few types of tags, kind of like ACLs and file attributes. Deny tag, Rating tag, Subject tag, Alert tag, Emergency tag, Meta tag, etc...

    rate:88/100
    subject:spring planting

Then BlueSky can set client defaults to allow Alert tag usage by anyone, but the Emergeny tag only by major news orgs? or governments? These would be effective because people generally accept defaults. Rating tags would help standardize rating systems and allow evaluation of a user by what they recommend or don't recommend. Subject tags would build community concensus on the tag name/tag text because people want their stories or comments to have the popular tags. Abuse of popular tags would also help people judge and filter users and other sources.

Tag types could also help limit downloads by processing Emergency first, then Deny tags, then rating tags, etc...

All this talk of user-defined moderation is great, however this is fundamentally incompatible with content that is legal in one jurisdiction, but is illegal in another, and thus must never be served in the first place to users of the second jurisdiction.