In a little while, LLM may enable "mute this discourse" functionality where you describe the particular bullshit you don't want to see and it filters out more precisely what you want than filtering by keywords or whatever.