I have an app that does Hacker News with AI; it analyses all the stories and comments for a number of criteria, and tags them so you can skip stuff you don't want to see. I should get around to actually publishing it but I've been lazy.

One of the fun things I noticed is the psychological impact of framing. A comment that might've made you feel the need to reply before has less emotional weight if it's highlighted in red and a diminished font. Same thing for stories; if you would normally disagree with a story and it would make you want to comment, you feel less like commenting if the story is rated as 'lacking evidence', 'unsupported by research', 'personal anecdotes only', etc. It drives down the feeling of needing to engage. Which is horrible for site engagement, but good for mental health (I think).