So what you are saying is that if youtube cannot solve the problem of scale, those problems should be dumped on the general public instead of youtube decreasing their scale? That is some interesting logic.
I am actually all for YouTube becoming less of the behemoth in the market. One way of accomplishing that would be to enforce human moderation instead of machine moderation. Because nothing at the scale of YouTube can (sustainably) use mostly-human moderation and I suspect even "machine-assisted human moderation" would simply require too many people.
If we take "you need about 3x the time of a video to make a considered moderation decision" as a baseline, and trust the numbers from https://merchdope.com/youtube-stats/ as valid, we would need....
300 hours per minute * 1440 minutes per day * 3 moderator hours per day. Let's round that up to 1.3 million hours of daily reviewer time. Let us also assume we have super-human reviewers that can squeeze out 8 hours of solid reviewing a day. That means we need 160k - 170k reviewers. And then we need to account for illness and hols, so make that about a quarter of a million people, to keep up with the incoming and maybe make some inroads on what's been uploaded in the past.
That is actually less than what I expected, I thought the numbers would end up in the "low millions" (call it 3.14 million).
It's the cleverest trick the social media companies pulled. "It's too hard!" So they don't do it!
I'm imagining a world where the answer was "tough luck, you have to or you don't get to do business". Imagine a Facebook where due to the enforced human moderation you were only able to make a single post a day. That actually sounds like it might solve a lot of problems!
If we take "you need about 3x the time of a video to make a considered moderation decision" as a baseline, and trust the numbers from https://merchdope.com/youtube-stats/ as valid, we would need....
300 hours per minute * 1440 minutes per day * 3 moderator hours per day. Let's round that up to 1.3 million hours of daily reviewer time. Let us also assume we have super-human reviewers that can squeeze out 8 hours of solid reviewing a day. That means we need 160k - 170k reviewers. And then we need to account for illness and hols, so make that about a quarter of a million people, to keep up with the incoming and maybe make some inroads on what's been uploaded in the past.
That is actually less than what I expected, I thought the numbers would end up in the "low millions" (call it 3.14 million).