It's despicable, sure, but not at all surprising.
It is natural to try to see patterns in the moderation, but I think the largest pattern is that there is little pattern. They instead have an job that is impossible to do at the scale they do it at the budget they do it at. They therefore make a lot of mistakes. They hate to admit these mistakes so they are constantly trying to defend the indefensible rather than admit that there is no way to have a uniform moderation policy that works on the hundreds of hours of video that are uploaded every minute.
Yes, and it drives me bonkers because everyone who makes these kinds of complains tend to like the narrative that YouTube is 100% hellbend on just being bad because Google Is Bad (tm)
I'd be curious to know what the strikes were for, since they never discuss them. But non-political channels get burned by unfair strikes all the time, including on very old videos.
Another possibility is the use of targeted reporting by fans of those featured in videos.
This is the money shot right here. They can’t admit the impossibility because if they did they further open the door to the private platform versus town square debate (which doesn’t seem to have certain consequences either way… yet)
In this case, they're putting profit over ethics leading to negative effects, so this one doesn't. Furthermore, Hanlon's razor is a false dichotomy: In this case the result is unethical yet neither malicious nor stupid.
Do they really though? Is there evidence that there are *any* effective strategies for reducing viewership of mind-rotting extremist content at scale?
It is my understanding that what the right wing is going through in 2021 is kinda what Muslims have gone through in 2016, with lots of outrage because some visible targets have been taken down, the malicious actors attempting to fan the outrage and get the entire community to protect the actual evil content, the feeling of being singled out for termination, etc.
How much overlap do you think there is between those who watch conspiracy videos and those who watch the videos that debunk them? I'd guess it's really small, probably <5% and possibly <1%—depending on how much Youtube's algorithm recommends the latter to the former. Actually, if they wanted to maximize view counts, they could tweak the algorithm and keep both videos, thus both keeping the gullible audience and driving outrage views from the skeptical audience.