I did some superficial study on what makes social media terrible, a few years ago. I’m surprised how close the conversation came very close to the elephant in the room without actually getting there: Eliminate advertising on the platforms.
I could pile on analogous examples from television (which might depend on age and being in the United States), but prior to cable’s dominance, it wasn’t hard to untangle a winning marketing strategy. You want ads in front of gullible people. Emotional engagement makes people less skeptical. The easiest emotions to consistently provoke in people are fear and anger. Therefore, each act break should be an upsetting surprise, so that the audience won’t change the channel and might focus on the commercials.
Social media accelerates and optimizes that feedback loop to drive up ad revenue. So, not only is the landlord likely to miss your request to quiet down the belligerent creep, but it’s actually in the landlord’s best interests to keep the creep and boot you out for drawing attention to his belligerence. It’s in the platform’s financial interest to gamify your retweet of something that’s going to make people angry and, if you don’t, the algorithm will get it in front of someone who will.
Basically, if any of these companies got rid of advertising, that outrage-feedback cycle falls apart, and the algorithms start to align themselves with showing you things that are actually relevant, instead of provocative. Yes, there are exceptional cesspools whose entire purpose is to radicalize people, but most of them struggle to stay relevant beyond the occasional investigation after an act of terrorism to find the kid’s manifesto, so they don’t really count.
If that’s not viable, there are a couple of other features that I’ve saw when I surveyed the Free Software social networks that seem to mitigate problems.
- Onboarding is crucial to establishing a healthy culture. Explain the features so that they don’t get misused and annoy people. Explain the moderation rules. Explain the algorithm. Help craft an introductory post to get people started.
- Mastodon’s “content warning” system means that people need to take an additional, affirmative step, if they want to be annoyed, though using them well requires training.
- Kill infinite scrolling. People are going to want to read to the end of the page, so there had best be an end to the page, unless the goal is to train them that they’re not allowed to leave.
- Have personal moderation tools beyond a “dead to me” button. For example, Scuttlebutt distinguishes between a response that’s part of the threaded conversation and a response that’s it’s own thread with a reference, and it’d be nice if that decision could be retroactive for the person being replied to, to say “that’s off-topic.”
Without more information about the inner workings of the systems or the people actively using the system, those would be my targets.
