There is one solution that is not yet economically feasible. It is practically possible to train a “guardian” ML model to perform vetoing of notes that reach the person’s eyes, and have a tunable threshold 🐶🐾🤔

I had no idea. Perhaps a solution is at hand.

imagine going to a party where everyone is kind and and friendly, except that one guy who is annoying, creeps on you and makes you feel uncomfortable. heck, he may even hate you and want to plot your death and the death of anyone who is remotely like you. now, imagine going to a party but every person is that guy. This person not only internally thinks you're a shitstain but openly vocalizes their opinions as facts. And it's a whole room of these people, all cursing you in chorus for simply being who you are. Oh, but there's one person who is kind and nice. Doesn't seem worth it to stay, does it.

I think this allegory is a smidgen hyperbolic, and entirely rhetorical but I'll assume it's being asked in good faith: No of course I wouldn't tolerate that environment.