Tag: facebook

For Whom the Search Trolls

Photo by Ross Williamson

One of my main topics of interest with regard to the Internet is the notion of what my friend, the writer Jeff Turrentine, calls “epistemic closure.” Let’s face it:  when it comes to information, it’s all too easy to find evidence out there for just about any bias or belief we can name; and I am far from the only person to ask what effect this has on our political process.

It seems self-evident that an environment like Facebook is generally an echo chamber when it comes to socio-political issues, and I do believe this plays a role in maintaining, if not increasing, balkanization.  After all, it’s hard to find a more potent ideological brew than a peer group armed with quips and clips that favor one’s established politics.  Additionally, social media tends to increase the number of headlines we see without necessarily increasing the volume of in-depth reporting we read.  While this may not matter much in a macro view (i.e. whether we’ll vote democrat or republican), it does matter a lot more in the day-to-day micro complexities of governance; and I would not be surprised if the 140-character attention span we’re fostering aggravates the tendency to adopt associative political positions. The fact that any given issue can generally be placed in either a blue or red column is not necessarily good for us citizens, but it is a boon to most marketers, especially now that news and entertainment have irrevocably mated to produce a mutant child as yet unnamed. And that brings us to the matter of search engines.

This video from the founders of an alternative search tool called DuckDuckGo touts a small study they’ve done indicating that Google’s personalized search can have a negative impact on our democratic process, precisely by providing the aforementioned epistemic closure. In other words, their initial research shows that Google has enough data about each of us to tailor results on a polarizing term, say abortion, to deliver what we most likely want to find.  Of course, DuckDuckGo has something to sell, but that doesn’t make the question they raise invalid.

This article by Gregory Ferenstein at TechCrunch addresses the issue dispassionately, concluding that more research is required to determine whether personalized search really has any effect on people choosing to seek out information they need, regardless of whether or not it’s what they want to hear. Scientifically, I’d have to agree with Ferenstein; but anecdotally, my instincts lean toward the hypothesis offered by DuckDuckGo.  Multiple times a day, both conservative and liberal friends post articles from news aggregators that sound just a little too spot-on to be taken at face value; and in fact many of these stories are full of holes and editorial hyperbole.  Stepping back and watching the posts roll by, I am reminded to consider the question of who benefits from all these collisions that seem to cancel one another out like particles and anti-particles.

And so, the big-picture concern is this:  a very tiny consortium of corporations, much smaller than the consolidated media conglomerates, own the revenue streams generated by our online activity. In fact, for now, one corporation owns almost all of search and ad service on the Web. So, if it is in the interest of advertisers to narrow rather than broaden our paths through cyberspace, and this winnowing can be made to look like a service to us users, are we in danger of having our perspectives constricted while being sold the promise of limitless access?

Keep in mind that as users we may want the world at our fingertips but that the brass ring for marketers is the targeted advertisement.  While there’s no question that a search for a local merchant or restaurant is more convenient when Google uses contextual data to second-guess what I’m looking for, there are other circumstances in which sorting based on my profile feels just a tad invasive and manipulative.

Why isn’t the Internet breaking?

During the squabble over SOPA and PIPA, one of the underlying (and possibly just lying) PR bullets coming out of Silicon Valley was that the actions called for in the bills would “break the Internet.”  And when that wasn’t the claim, the most consistent complaint was that the bills would chill free speech.  But in the wake of violent protests to an online video that may be related to the deaths of American diplomats, it turns out there is suddenly room for discussion about both speech and algorithmic solutions to thorny problems in an otherwise “free and open Internet.”

According to this piece by Somini Sengupta in yesterday’s New York Times, there is not only room for discussion, but it seems we’ll be having this discussion for quite some time and hearing from many parties. If nothing else, this article makes plain that the concept of free speech is no more universal inside the conference rooms at Google and Facebook than it is among nations.  So, if speech is relative, and algorithms can theoretically be written to correspond to definitions of “hate speech,” what was all that flap about SOPA and PIPA? I know the mechanisms requested by those anti-piracy bills are different from those required to address the issues cited here, and I don’t know anything about writing code; but it seems to me that the math problem in the case of analyzing hate speech worldwide is a hell of a lot harder than cutting off funding sources to a finite number of torrent sites.

The irony, of course, is that the makers of The Innocence of Muslims are fully protected by the First Amendment, even though what they chose to do with that right is irresponsible and loathsome.  And even if someone did manage to come up with a universal definition for hate speech, and a programmer managed to write code to seek it out on the Web, it’s possible this film that started all the trouble might not even meet the narrow criteria that would need to be written.

By contrast, the transactions made possible through torrent sites are not protected by the First Amendment, are comparatively easy to define, and extremely easy to locate.  Yet, we were led to believe that neither law nor technology could possibly have stopped or even mitigated piracy.  I don’t know.  As I say, I don’t write code, but it sounds a little bit like the masters of the Internet are clearly capable of flying F-18s when they need to but suddenly incapable of driving golf carts when they don’t want to.