Can’t be that hard with AI. Just don’t let the send tweet button work on combos of certain words.
It isn't. In fact, I've personally seen a demo by a start-up a few years ago that did just that but went quite a lot further and also constantly mined social media trends to be able to recognise e.g. the common dog whistles of the day, could detect sarcasm reasonably well, etc.
And that was two or three years ago, i.e. in the technological stone age from today's perspective.
No idea what happened to them but they've hardly had the victory march they could have had with that product of theirs, now, have they?
The technology is there. In fact, it wouldn't even be strictly needed as long as human moderators exist.
It's the willingness of SM companies that's the issue.
And, yes, I've been reading along. Absolutely vile! I'm all for free speech and healthy debate - but that's entirely possible without resorting to vile threats!