That also happens on every platform. Algorithms promote posts with a lot of engagement. Now, what is the first thing you do when you see a bad post? You want to tell the whole village about how bad it is and then clown on the guy who made it. But, in doing so, you just gave them shares, views, and comments, which the algorithm rewards all the same.

I'm beginning to think that social media algorithms are some of the first prominent examples of harm caused by misaligned AI systems. These systems, which were supposed to help people find content that would interest them, became a game of min-maxing for variables, subtly changing both the behaviors of creators and consumers. As a result, that innocuous goal evolved into a system which facilitates self-reinforcing filter bubbles, the rise of the rage economy, the influencer meta, and much of the slop you see on a daily basis... All while the very companies that implemented them now have a perverse incentive to keep those systems because they suffer from a bad case of Goodhart's Law: the algorithm maximized for quantifiable variables (engagement, views, etc) that it made nearly every social site hell to use.