Fediverse

If implemented poorly, yes, but no worse than existing practices.

e.g. one example of democratization is placing labels next to usernames. Knowing that someone has a label for "account <2 weeks old" or "posts excessively more (I dunno, 10x?) than comments - might be an unregistered bot account?", "contentious user receiving >10x more downvotes than upvotes" etc. - btw the labels are icons, not this long text explanation that I just made up here:-), places the power in the hands of the user to decide what to do with this information. e.g. I might go ahead and respond to the contentious user, or I might not, but if I do perhaps I'd really solidify my points rather than presume a friendly audience that will make allowances for a slackened response.

Btw you can also attach personal icons to usernames as well, which helps you notice whatever feature that you want to call out as you scroll through large comment sections. I do it for friends, but ofc someone could just as readily do it for the opposite:-).

Another example, probably much better and more relevant, is how comments are automatically collapsed or automatically hidden (two different procedures, albeit related to one another, allowing different threshold values for each) (edit: whoops, I forgot to say that both of these are based on the ratio of up vs. downvotes). If the user decides that they want to see only friendly, nice, and casual content then... so be it, they can ask for such - essentially relying on community consensus to make that decision for them. Mind you, I personally don't like either of these and so have turned them off, and that's okay too, yet I still respect so much that they were provided.

Another quick example: during the signup wizard, or anytime afterwards, the user can ask to filter content containing keywords such as "Trump" or "Musk", choosing between total removal, partial to allow a little bit through, or no filtering at all. Again I don't personally use these options but... it really is so nice to have been offered choices!

Now before you point out how subversive actors can game the system - which... they can, just as now - I want to point out why I call this "democratization". Right now, a moderator gets / has to decide such things, which affect all of the community members at once. Is political content allowed? Is someone being an unfriendly dick? Should a minimum community-specific karma score be acquired (from responding to existing posts, or voting) before they are allowed to make a post? All of these are things that moderators can / must decide on, whereas with even just these approaches that I mentioned above (and there are many more that I left out) that is no longer required. With these tools, while the old way still works and a mod can remove something, now they don't have to, and can leave the choice up to the user. And for things like a label or a mere collapsing of a comment, it is readily reversible to uncollapse it and read its content if desired in some context. Whatever the user wants is what drives that interaction.

Notice then how it changes the moderator dynamics: whereas before there was only a binary set of options, to either remove the content entirely or to not, possibly locking it or just leaving it alone, but since I want to squeeze in the phrase that it's a "binary decision" I'll say it like the mod decides to either take action or to not. Well now, new options are available, where multiple types of actions could be taken, but all of them driven by the user, rather than the moderator. This shifts the power away from the authoritarian controllers into the hands of the common people.

Mind you, it's a separate issue whether democracy itself is a good thing, in comparison to authoritarianism, but in any case, PieFed does provide democratization of moderation, so at least that seems an accurate way to describe the situation, imho?

And there are tweaks that definitely should be done, which the developers are very much aware of. e.g. for things like auto-collapsing of content, allow votes only from "trusted" locations (whitelist), or allow from any location unless it is "untrusted" (blacklist), or just allow all, or maybe not paint with such broad brush strokes and instead dig deeper e.g. allow only votes from community members (which itself should be investigated: if someone "joins" but never actually comments or posts, are they truly a "member" then, for that purpose? or are they rather a bot that is trying to pretend like a human for manipulation purposes?). Obviously we cannot let perfection be treated as the enemy of good, and PieFed is at least making strides towards what I see as a good aim. Lemmy in contrast is not - allowing anyone at all who wants to spin up an instance and start manipulating voting to do so - hence at a bare minimum I think it is exciting to see PieFed doing more than Lemmy there (even if someone wants to quibble over the specifics, which ngl has merit as well, but is a separate thing than whether the overall goal is a good one).