Enough of them to give rise to the term "algospeak" which means using words like "unalive" in place of "kill" to avoid automated censorship.
https://en.wikipedia.org/wiki/Algospeak
Meanwhile you can report a bot who's posted 20+ comments under a video to advertise illegal drugs and all of the reports and subsequent appeals will consistently come back as "No violation found".
> you can't say certain words in the video or mention them in the title or you get demonetized
That's not censorship problems.
That's advertisement problems. That's conflicts of interest problems. That's incentives problems. That's people-who-post-videos-just-to-make-money problems.
Well, okay, it can easily be turned into censorship problems: instead of just demonetizing the video, don't show it to anyone. It's quite a fine line, but the line is indeed there.
The Chinese one is to do with China, the American one is to do with America.
Does "kill" have some type of salient political valence that I'm not aware of?
This seems like a fairly blunt attempt at quality-of-life improvement for the general platform vibes, no? Put some friction on the (legitimate) nutjobs who just want to say "Kill X, kill Y" all the time and are so insane they can't figure out euphemisms?
This has been happening for 10+ years on e.g. YouTube, you can't say certain words in the video or mention them in the title or you get demonetized. Nothing to do with China.