Destroying creativity and incomes

[Excerpted and edited for brevity from Cory Doctorow’s book The Internet Con: How to Seize the Means of Computation]

Google created Content ID, an "audio fingerprinting" tool that was pitched as a way for rights holders to block, or monetize, the use of their copyrighted works by third parties. YouTube allowed large (at first) rightsholders to upload their catalogs to a blocklist, and then scanned all user uploads to check whether any of their audio matched a "claimed" clip. Once Content ID determined that a user was attempting to post a copyrighted work without permission from its rightsholder, it consulted a database to determine the rights holder's preference. Some rights holders blocked uploads containing audio that matched theirs; others opted to take the ad revenue generated by that video. One big problem with this is the inability of Content ID to determine whether use of someone else's copyright constitutes "fair use”, which is legally permitted. A fair use determination, unfortunately not easy for a program to do. Computers can't sort fair use from infringement. That means that filters block all kinds of legitimate creative work, especially work that makes use of samples or Quotations.

A lot of creative work is similar to other creative work. Part of Katy Perry's 2013 song "Dark Horse" is effectively identical to a part of "Joyful Noise," a 2008 song by a less well-known Christian rapper called Flame. Flame accused Perry of violating his copyright. Perry eventually prevailed. But YouTube's filters struggle to distinguish Perry's six-note phrase from Flame's. Filters routinely hallucinate copyright infringements where none exist - and this is by design.

It's pretty trivial to write a filter that blocks exact matches. But an uploader could get around a filter by just compressing the audio ever-so-slightly, below the threshold of human perception, or cut a hundredth of a second off the beginning or end of the track, and this new file would no longer match. YouTube filters employ "fuzzy" matching. They block stuff that's similar to those things that rights holders have claimed. Rights holders want the matches to be as loose as possible, because they would want to stop someone from getting even a mildly altered song for free. Unfortunately, the looser the matching, the more false positives. Don’t even try to earn money from your classical music performances, because any performance of Bach, Beethoven and Mozart will sound too much like Sony Music recordings. Even teaching classical music has become a minefield and you’re likely never going to earn any money from it on YouTube.