Bottom dwellers
This week Google announced it would downgrade in its search results sites with an exceptionally high number of valid copyright notices filed against them. As the EFF points out, the details of exactly how this will work are scarce and there is likely to be a big, big problem with false positives - that is, sites that are downgraded unfairly. You have only to look at the recent authorial pile-on that took down the legitimate ebook lending site LendInk for what can happen when someone gets hold of the wrong side of the copyright stick.
Unless we know how the inclusion of Google's copyright notice stats will work, how do we know what will be affected, how, and for how long? There is no transparency to let a site know what's happening to it, and no appeals process. Given the many abuses of the Digital Millennium Copyright Act, under which such copyright notices are issued, it's hard to know how fair such a system will be. Though, granted: the company could have simply done it and not told us. How would we know?
The timing of this move is interesting because it comes only a few months after Google began advocating for the notion that search engine results are, like newspaper editorial matter, a form of free speech under the First Amendment. The company went as far as to commission the legal scholar Eugene Volokh to write a white paper outlining the legal arguments. These basically revolve around the idea that a search algorithm is merely a new form of editorial judgment; Google returns search results in the order in which, in its opinion, they will be most helpful to users.
In response, Tim Wu, author of The Master Switch, argued in the New York Times that conceding the right of free speech to computerized decisions brings serious problems with it in the long run. Supposing, for example, that antitrust authorities want to regulate Google to ensure that it doesn't use its dominance in search to unfairly advantage its other online properties - YouTube, Google Books, Google Maps, and so on. If search results are free speech, that type of regulation becomes unconstitutional. On BoingBoing, Cory Doctorow responded that one should regulate the bad speech without denying it is speech. Earlier, in the Guardian Doctorow argued that Google's best gambit was making the argument about editorial integrity; publications make esthetic judgments, but Google famously loves to live by numbers.
This part of the argument is one that we're going to be seeing a lot of over the next few decades, because it boils down to this bit of Philip K. Dick territory: should machines programmed by humans have free speech rights? And if so, under what circumstances? If Google search results are free speech, is the same true of the output of credit-scoring algorithms or speed cameras? A magazine editor can, if asked, explain the reasoning process by which material was commissioned for, placed in, or rejected by her magazine; Google is notoriously secretive about the workings of its algorithms. We do not even know the criteria Google uses to judge the quality of its search results.
These are all questions we're going to have to answer as a society; and they are questions that may be answered very differently in countries without a First Amendment. My own first inclination is to require some kind of transparency in return: for every generation of separation between human and result, there must be an additional layer of explanation detailing how the system is supposed to work. The more people the results affect, the bigger the requirement for transparency. Something like that.
The more immediate question, of course, is, whether Google's move will have an impact on curbing unauthorized file-sharing. My guess is not that much; few file-sharers of my acquaintance use Google for the purpose of finding files to download.
Yet, in an otherwise sensible piece about the sentencing of Surfthechannel.com owner Anton Vickerman to four years in prison in the Guardian, Dan Sabbagh winds up praising Google's decision with a bunch of errors. First of all, he blames the music industry's problems on mistakes "such as failing to introduce copy protection". As the rest of us know, the music industry only finally dropped copy protection in 2009 - because consumers hate it. Arguably, copy protection delayed the adoption of legal, paid services by years. He also calls the decision to sell all-you-can-eat subscriptions to music back catalogues a mistake; on what grounds is not made clear.
Finally, he argues, "Had Google [relegated pirate sites' results] a decade ago, it might not have been worthwhile for Vickerman to set up his site at all."
Ten years ago? In 2002, Napster had been gone for less than a year. Gnutella and BitTorrent were measuring their age in months. iTunes was a year old. The Pirate Bay wouldn't exist for some months more. Google was two years away from going public. The mistake then wasn't downgrading sites oft accused of copyright infringement. The mistake then was not building legal, paid downloading services and getting them up and running as fast as possible.
Wendy M. Grossman's Web site has an extensive archive of her books, articles, and music, and an archive of all the earlier columns in this series.