I'm glad you're looking into working on this! In my opinion, blacklist providers (and their integration into relays and clients) should be as conservative as possible (limited to illegal content, or inline with a special-purpose relay's mission). For other type of "unsavory" content, you should look into NIP-32, which was built to support multiple ontologies for content classification and rating. I think that will be a much better solution for web-of-trust content recommendations and soft moderation.
I’ve been thinking deeply about content moderation on Nostr.
I’m very pro-freedom-tech but there are problems we will need to solve to scale and be acceptable to a wider audience of user.
To help get the ball rolling on improving Nostr’s ability to keep bad actors from poisoning the well here’s a proposal for a freedom maximising strategy for dealing with illegal or just insanity content (as determined by Nostr users)
I know this’ll be controversial so I’m looking for feedback and refinements before trying to build something to get this ball rolling.
https://gregwhite.blog/nostr-content-moderation/
Cc nostr:npub1yye4qu6qrgcsejghnl36wl5kvecsel0kxr0ass8ewtqc8gjykxkssdhmd0
Discussion
Yessir I’m very on board with that.
I think since this strategy is entirely consent based (every level of Nostr user chooses what blocklists they want to use) most relay providers will be conservative as you mention.
But I’m open to any ideas on making that as explicit as possible.
Thanks for reading and giving feedback!
The lists a relay applies has implications for its users — this might be a service, but is more likely to be a hindrance, because it's easier to apply more filters, but not so easy to un-filter results.
Using lists makes sense, but applying mute lists is probably not a good idea, since you can mute something for multiple different reasons, which is why I mention NIP 32, it gives you the granularity you need to make a more educated filtering decision.
That’s a good point I’ll keep digging into that.
For relays I think the advantage of subscribing to lists is they have more tools to not get taken down by law enforcement nits not about deciding what users should see, it’s just about being defensive against the statists until the statists can be overpowered by freedom tech.
And core to this is user choice. Users should choose relays that manage content the way they agree with (or need to because of local laws). Including no management of content at all!