Comment by Andre607 | Hacker News Reader

Comment by Andre607 | Hacker News Reader

YouTube's approach to takedowns is, simply put, appalling. I'm not just referring to the automated Content ID system, which has a vast catalog of false identification, but am talking about the manual, human confirmation of takedowns.

My 'favourite' (in the sense of most egregious) example of what I am talking about is the case where YouTube's Content ID flagged the sound of birds in someone's garden as being copyright infringing [1]. So far, par the course for a mistaken automatic identification. But the outrage comes after that: the copyright claimant (the notorious Rumblefish) reviewed the claim and confirmed that it was valid! In other words an actual human being looked at the video of someone in their garden, and confirmed the claim that the sound of the birds in the background was copyrighted. Actions like this are indefensible and highlight the outrageous monstrosity of YouTube's takedown system - -beyond Kafkaesque in its total disenfranchisement of users.

[1] https://www.techdirt.com/articles/20120227/00152917884/guy-g...

Similarly, a friend of mine recorded waves crashing at a sea and it got flagged by Content ID.

Some music video (by Sony subsidiary) contained a small sample of waves crashing from a completely different sea.

I told him to appeal and appeal he did and yes the claim was upheld.

I told him to fight on because this first level upheld is what throws people off and what the Content ID abusers are hoping for.

On the 2nd appeal they have to have someone with legal authority to really pursue the claim.

That is no "real" human will look at the case on its merits until 2nd instance.

Wow! How did this story end?

> In other words an actual human being looked at the video of someone in their garden, and confirmed the claim that the sound of the birds in the background was copyrighted.

i think it is likely that this human is just someone employed from mechanical turk who gets paid on the basis of how many videos they process in an hour and not actually an employee of rumblefish

In this case the CEO of Rumblefish, Paul Anthony, confirmed that it was in fact one of their employees [1]:

> They still work at RF, feel absolutely terrible about what happened and are taking a whole new approach to claim/dispute reviews. A mistake like this in the hands of the right employee is a game changer and brings about significant improvement. In the hands of a less-competent employee...they fold and get worse and that isn't acceptable for me. I hope and trust that it's the former. I believe that although it was a terrible oversight, that it was an honest mistake by this individual.

[1] https://www.reddit.com/r/IAmA/comments/q7via/im_the_ceo_of_r...

Note that since that particular incident, Rumblefish has had continued incidents of blatantly erroneous coypright claims - like claiming ownership of a US Navy rendition of "America the Beautiful" (which is public domain) and proceeded to monetize it! [2]

[2] https://boingboing.net/2015/07/03/july-4-rumblefish-claims-t...

Rumblefish of course describes these incidents as "honest mistakes", which completely ignores the inconvenience they cause people and the fact that they have to make noise via social media or tech news outlets for them to take action - the real concern is how often this happens that we don't hear about.

Reply on news.ycombinator.com