@dannekrose @futurebird
Some folks are trying to work on that. The problem is that there has been a conflation between ‘you control what is in your feed’ and ‘you don’t see anything that you didn’t put in your feed’.
My primary objection to platforms like Facebook and Twitter is that they are psychological profiling tools that sell ads / propaganda and masquerade as communication tools. Federation kind-of helps that, but Google showed with GMail that it’s possible to capture a sufficiently large chunk of a communication graph that you can still build these things. The main benefit of the Fediverse in that respect is breaking the monopoly: it’s as easy for Google or Facebook to crawl the Fediverse and build profile information as it is for anyone else, but they can’t augment it with private message content.
The problem that Mastodon and friends try to address is the lack of agency of users. Twitter turned the web back into TV: you could change channels, but they still controlled the content that you saw. To address this, they started at the exact opposite end: you see posts from people and hashtags you follow, but absolutely nothing else. The choices are all digital: follow or block people / hashtags.
In reality, people want much more analogue controls. Show me more cats (but not every post about cats). Tune down the US Politics (but a few posts a day is okay). Show me a few random posts that are popular.
Completely removing a recommendation algorithm gave a nice clean slate to start from, but that’s all. Recommendation algorithms are useful. The problem with other platforms is not that they have recommendation algorithms, it’s that their algorithms are opaque and under someone else’s control. The goal for people working on such things here is that the recommendation algorithm must be completely under the user’s control and must be able to answer, for any post that they see, why they see it. Oh, and it must actually be useful. It’s coming, but building it is hard!