Over reliance on algorithms has degraded the user experience to the point that the average user is drowning in ragebait and extremist politics, because they drive up engagement. Just like a toddler, algorithms don’t discriminate between good and bad attention, so everything that gets clicks is thrust forward. Now, you could hope to train the algorithm to show you only postive things, but engagement is engagement and the algorithm curators often engage in rage farming, where your feed is injected with things that are likely to enrage you.
You can avoid this by installing an RSS reader, going to your favorite sites, and manually adding a RSS feed. Now, your reader has things that you manually selected, with the added bonus of having a content pipe free of malicious interference. You can also divide topics in a way that you can avoid certain themes and news until you decide to engage them.
I’ve actually created a Lemmy instance (Lemmy.link) to bridge the gap between RSS feeds and Lemmy. We have 36 communities (topics) you can subscribe to from your home instance. Each community has the RSS sources listed in the “Feeds” section of the community sidebar. I’m always open to feedback on new community ideas or how to make it better. Currently I’m working on having the bot parse the article and ignore if it is an ad or sponsored and also including the channel name in the subject line for YouTube sources. Eventually I’d also like to train a model on which topics don’t receive upvotes or comments and attempt to ignore them for a better signal:noise ratio.
Was planning on doing something similar. I think adding more topics like science and specific sports could be useful.
Edit: My scrolling cut off half the list of communities. Looks like you already have those.