I’ve been meaning to make git issues, but in a nutshell, one is “community taxononony”
Instead of being pigeon-holed into a single community, every community would be part of an inherited hierarchy, like a class system in programming. /c/thelastairbender might be part of /c/animation, or /c/television; perhaps both?
Organization would be mutual. Moderators of each have to approve to join and remain in the hierarchy, though the “initial structure” of the community could be set up by admins I suppose. The sub community inherits “global” rules from their parent communities, but can have their own rules as well.
And what’s the point of all this, you ask? Well, way I see it, Lemmy has a “niche” discoverability/attention issue, where big engaging communities like politics crowd out smaller niches. But being a sub community would show all its posts in the communities up the hierarchy as well, getting them the visibility of a “big” community while remaining in the niche. It would allow focused communities to exist, but users browsing bigger communities to see them as an appropriate topical thing. This aggregation is user configurable, of course, but I think it’s very important that this visibility be the default.
And in terms of programming, I think it would be feasible? Admittedly I don’t know the architecture, but it seems like it would fit with existing paradigms.
Another idea I have is a replica of Twitter’s “community notes” feature. Perhaps if a comment gets enough upvotes and is flagged by the comment writer as a “community correction,” and fits certain criteria (like being below a word count, maybe a certain percentage of upvotes being from the host instance), it’s automatically displayed below the original post’s title.
This would allow, for example, clickbait or questionable sources to be called out, or misleading titles to be clarified. Or perhaps the source of original reporting can be hyperlinked.
Theoretically this is a mod’s job, but I feel that:
-
Mods don’t want to be heavy-handed
-
They’re often overworked/short on time.
-
And frankly, let a lot of clickbait/ragebait posts slide anyway.
And for all of Twitter’s failures, this particular feature is a good idea.
Again, it ties into the idea of “attention control,” to try and give information hygiene a chance over people’s impulses.
Mind you, these are very rough ideas. They probably need to be peeled apart, but I do feel strongly about the gist of what they are trying to correct.
Human brains just aren’t wired for citations. Especially outside academia I guess.
I think it would help if people were more “LLM literate” though, eg they took a lesson in school on how they work at a low level. Folks would be horrified they ever put so much trust in them.