YouTube takes down the Ig Nobel show because of a 1914 recording

Assuming there is no current legal rights holder for this recording, I firmly believe that companies that falsely assert their rights should be held criminally liable for theft.

There is no such thing. Data is just bits. Really big numbers. Asserting ownership over numbers is simply delusional. The second that number is published, it's already over.

Non-fungible tokens do absolutely nothing to change those facts. They just let people delude themselves into believing they actually own stuff. The only thing they own is the token.

I don't know the full details, but as a point of interest, the Ig Nobel awards was a project at MIT, and after a dispute between the editor, Marc Abrahams, and MIT, he claimed to own the intellectual property rights and MIT dropped it and he moved it to Harvard where he is an alum.

here from "The Tech" http://tech.mit.edu/V115/N48/ignobel.48n.html

"Legal rift takes awards from MIT

For the past four years, the Ig Nobel prizes were awarded at MIT, by the Journal of Irreproducible Results and its successor, the Annals of Improbable Research. However, a legal dispute that arose between the MIT Museum - the publisher of AIR - and its editor, Marc Abrahams, caused the ceremony to be moved from MIT to Harvard.

Abrahams and the MIT Museum produced AIR for a year without a contract between them, but the museum wanted to create another organization to publish the magazine because handling the AIR required too much effort from the museum staff, said Warren A. Seamans, director of the MIT Museum.

Last March, contract negotiations broke down, and Abrahams claimed sole control of AIR. To avoid a lawsuit, MIT abandoned the magazine and the Ig Nobel prizes."

1. The viewer or listener of the music, you might or might not get the right to listen to music in your "territory" or country, and often the music is monetised which means you get ads, or a portion of your subscription revenue is apportioned to the "view";

2. YouTube itself, whom decides on the viewers' right to listen to the music based on their complicated set of algorithms or "claims" on the music track. They also have a complicated database of rights which includes not only rights holder relationships, but the much of the music catalogue itself;

3. A publisher of a piece of music. This copyright is for the composition (sheet music), and often YouTube may be able to work out the publishing right in a territory based on the melody match. So when you say "Content ID algo finds similarity in a dozen different recordings" this is in fact by design. Their melody matching system is in play as often publishers get paid on the composition;

4. The performance copyright of the music. This copyright is for the actual performance - whether it be live, on CD or MP3. Usually this is the record company, they will often upload or provide the music to YouTube and expect to get paid for performances. Content ID will then "claim" any third party copies of the performance and monetize them too (or sometimes block, depending on the territory of viewer, and what the publisher and performance owner wants to do for that territory);

5. Most countries also have music societies who collect on behalf of artists. So GEMA in Germany or PRS in the UK will collect money per view and by some complicated reporting will pay artists directly some small pittance every year if they are lucky. Some countries don't have societies which means YouTube doesn't need to pay. Where YouTube does not have an agreement in a country (less and less now) the music will not be monetized or may be blocked.

There are huge data reports which move around to keep this system going - and the collecting societies themselves work together to ensure that the money for a view of "Never Gonna Give Up" is passed to PRS who then occasionally will pay Rick Astley. It's a nice earner for a minority of musicians to have this income when all the cover versions and MP3 sales have dried up.

YouTube does have a lot of agreements with all the other parties and it's a constant job to renew and renegotiate these contracts constantly. You can see that in the rights notice you mention, this includes all the territories in which there are rights established: it's mind-boggling. No-one wants to lose control of this system and I imagine that makes it brittle and very difficult to disrupt...it's lawyers all the way down.

I hope that is not all too misleading!

So at present: owner tags some content as owned by them, content is fingerprinted, YouTube does similarity-matching, content that is similar gets flagged for takedown until the uploader intervenes. It's certainly not fair to hold the content owner accountable for false-positives any more than it's fair to hold the uploader.

(Now, in cases where the uploader pushes the "This content is legit" button and the content owner responds with the "no it isn't" button, I'd be very much in favor of incentives changing so an error at that point on the part of the content owner gets them raked over the coals. But there's no real legal mechanism for that to happen right now. Remember, this entire process exists as an alternative to the content owner's legal option: to sue YouTube or to sue individual YouTube users. YouTube doesn't want that).