In response to my post yesterday morning, I received a phone call from a Google spokesperson letting me know that my story (which was based on reporting by Danny Sullivan at Search Engine Land) was misleading. The representative had already contacted Sullivan who has posted this update to his original post.
Google is very sensitive to the idea that they are in any way giving YouTube special treatment in search results. They have very good, anti-trust reasons for caring about this, so I understand their concern.
The story that emerged from that conversation is both less sensational but possibly more interesting than what I originally wrote. Unlike what Sullivan described based on Google’s posts on the subject, that the “penalty was purely tied to number of notices acted upon,” it now seems possible that Google has kludged together an algorithm that somehow punishes the TorrentHounds and PirateBays, but not the YouTubes, Tumblrs, Twitters, IMdBs, etc.
Google is not divulging details of what are the hooks that attract piracy bosons to the file-sharing sites but let the user-generated content sites sail through to their normal search rankings, so we are left to wonder about the implications of this distinction. In order to do something significant against the wholesale file traffickers but not punish their own YouTube, Google has had to include YouTube in a “class” of sites that will not trigger the “signal” in the same way as the “bad” file sharing sites. Having nothing to do with the technical aspects of how this works (I leave that to Danny Sullivan, where he does a great—though sometimes overreaching—job of this) what it means for users is that Google is making a conceptual, almost moral, distinction between file-sharing sites and popular user generated content sites.
Reading the tea leaves, as I often do, I can think of two rationales (beyond protecting YouTube) for why Google might be trying to make this happen. First, consumer-grade user-generated content sites are more internally discoverable than file-sharing sites. So potentially, Google feels that more people are finding pirated material on file-sharing sites through search than they are on user-generated content sites where people tend to find things through social recommendation and well-exposed content navigation. The other point is that most of the popular user-generated content sites have very responsive take-down systems for removing infringing material, in stark contrast to the torrent sites.
Google is triangulating between the users, the sites and the copyright owners. They will say that the user’s interests come first, but since Google Search is a free service, it is probably disingenuous for the company to suggest that they do not take other stakeholders into account (especially ones tied to advertising revenue) in these kinds of machinations.
The mechanism through which Google is doing all of this is opaque (on purpose), but if all goes according to plan, it should achieve its objectives. Users will still find infringing material on user-generated content sites through recommendations and serendipity, to the extent that the sites allow it to be there, but users will have a harder time discovering that content through search. This will discourage the less hardcore file-sharers who will be easily redirected to legitimate items of equivalent interest at the top of the search rankings. Sites that have responsive takedown procedures (or whatever the distinguishing hook turns out to be) will not have their search rankings degraded, even if there is a large amount of illegal material posted on their site (i.e., YouTube). And the sites that primarily exist to share pirated material will get less organic search traffic through Google.
The point, I suppose, is that if user-generated content sites behave like responsible service providers, under the intent of the Digital Millennial Copyright Act, they should not be held responsible for the actions of their “rogue” users. The punishment is supposed to be meted out to the intentional infringers, the “bad actors.”
But, wait a minute. Are file-sharing sites and user-generated content sites really different “in kind”? Just because Google can insert some kind of wedge into an aspect of their functionality such that it can sort them programmatically, does that make them logically different? This may seem like a matter of semantics (which are, actually, quite important on the web) but it gets at what many users are confused about—and many advocates concerned about—Google’s search practices.
Search is not neutral. Google admits to having “200 signals” that they use to weigh rankings. These signals are all quantitative, but their inclusion and weighting in the overall search algorithm involve many judgement calls and conceptual distinctions. This is unavoidable. This is what makes Googling so effective. And it is what has allowed Google’s search experience to continue to improve even as it has negotiated more and more territory.
But when you are the biggest at anything, users expect a certain neutrality. And regulators expect it too. This is why Google was so concerned about the story I wrote yesterday, because it implied (inaccurately, as it turned out) that the company was giving preferential treatment to YouTube, which could be seen as monopolistic. In order not to single out YouTube, though, Google’s engineers have had to construct a class into which YouTube fits—that is distinct from the class that the evil file-sharers belong to. This is standard programming logic, but the rationality of it does conceal a very real preference for user-generated content sites.
In January, Rupert Murdoch, no paragon of corporate virtue himself, accused Google of being a “piracy leader,” in a tweet squarely aimed at YouTube. I don’t want to suggest that he speaks for all copyright owners, but the conflict here does suggest that the record labels, publishers and movie studios may not all see the same sharp distinction in flavors of piracy as Google does.
So, somehow, this little feature that Google has rolled out in order to regulate itself and be able to make a case against what it (and many, myself included) see as harmful SOPA-style government interference in the internet, points to the soul of Google. In order to do what they do, Google’s engineers have to make these distinctions, decide that some type of thing is more important or has more authority than some other thing. But in doing so they are engaging in cultural engineering on a massive scale.
Simply put, search results determine consumer and civic behavior. What we find becomes the basis of what we do. Shaping those factors is an awesome responsibility, and not one that I think Google takes lightly. But from time to time, that underlying power is exposed to public awareness. This, in a small way, is one such moment.
Không có nhận xét nào:
Đăng nhận xét