Where an unprivileged worker class has to go through hours and hours of video material of the upper societal levels in order to make sure that they satisfy the grotesque utopian "always happy no conflict" image of an all seeing censor, including the denial of their own existence.
I don’t have a Glassdoor account to confirm what this is saying but I think you’ll find at least circumstantial evidence here if you can view it. https://www.glassdoor.com/Reviews/Amazon-Data-Associate-Revi...
There are three major problems:
1. Audio, video and image recognition systems are nowhere near close to perfect and the upload volume means even a small error rate will result in a large amount getting through.
2. Users are constantly coming up with new ways to avoid the filtering systems making it a game of tag, but where one side is made up of tens of millions of people.
3. There isn't a central registry of who owns what type of copyright in each territory so there is no good way to know if the person uploading has the right to or not.
Then there are the issues of compositions... someone has the right to a music track, but only when synchronized to a particular video or someone inserting a frame of a copyrighted image inside their video.
And this will affect far more than YouTube. Copyright extends to all creative works including this comment. Image hosting services, forums, reddit, social networking sites, comment systems, review sites should all worry.
Who are you going to hire who can watch a video and instantly know every copyrighted piece used to produce it, judge whether or not the piece is one or another sort of "fair use" correctly/consistently etc., and be correct enough that YouTube feels safe taking on the corporate liability?
Getting this right is more than just "a human views the video once". Sure, there's some obvious piracy, but a loooot of things that are not obvious piracy would require active research.
They could catch the obvious stuff like people uploading Hollywood videos, but the reality is that humans will let tons of stuff through, and also accidentally block stuff that should have been allowed.
There's far more copyrighted content than a human can even view or listen to in a lifetime. There's just no way for a human to know what's in violation and what isn't.
People can host their own stuff. Simple. Don't host content you don't know the origin of and can't document you've got the right to do so.
How exactly would that solve the problem? Is there a step 2 where they also cut all the cables into and out of Europe?
Even if only 1% of videos get flagged for review, YouTube is still looking at ~50% of all it's employees working on reviewing video.
Of course, that don't work for copyright protection since there isn't a person alive who can recognize every copyright protected image/video/audio on the planet and whether the person uploading owns/licenses the copyright in every territory.
If this law is rolled out, I highly suspect that YouTube in the EU will consists solely of large channels that have a low rate of videos that violate policy.
400 hours a minute is 400x60x24x365=210,240,000 hours a year, equivalent to ~100,000 full-time employees. Anything less than $15/hour and they'll be hearing from Bernie Sanders, plus FICA and any employee benefits. Then they need real estate to work, utilities, parking, furniture, computer equipment, HR/IT/accounting support staff etc. The cost is in the ballpark for the estimates of YouTube's total revenue. Profits would be a fraction of that after all the existing expenses for hosting infrastructure and staff.
And the $100B in cash is across all of Google. They would have to be pretty dumb to spend it down continuing to operate something that could never be expected to turn a profit again.
[0]:Infantry-equivalent numbers are used.
https://www.theverge.com/2016/4/13/11387934/internet-moderat...
One might get someone unemployed and with no other choice to do it for a while, until another job comes by, but I don't really think anyone wants to do this for a living.