The Suit Magazine

Shortly after September 11, 2001, political activist Eli Pariser used the World Wide Web to send a message to anyone who would listen. The responses he received were rapid and overwhelming, attesting to the amazing connective capabilities of technology. But over the last ten years, Pariser has noticed some worrisome new trends. Now he's issuing a warning: if we're not careful, the internet may fall short of its once-great potential to enhance public interaction.

In an exclusive interview with The Suit, Pariser recalled the day he watched the live television broadcast of a grisly black cloud mushrooming over the World Trade Center towers before they fell to the ground. Stunned by the tragedy, he quickly went to work on his computer. "I knew I wanted to do something socially helpful," he said. A month after the attacks, he and a few friends posted a website calling for a multilateral approach to fighting terrorism. "And I was shocked," he said, "when 50,000 people signed our little petition."

That was only the beginning. Within a few days, Pariser was receiving countless calls that his site was crashing due to heavy traffic. Soon he was returning calls to major media outlets. "The phone was ringing off the hook. When the guy from the BBC called, I said it was me who crashed the website," he said.

In November of 2001, Pariser's website merged with MoveOn.org, an American nonprofit public policy advocacy group that opposed the war in Iraq and pushed for more progressive alternatives. Their efforts caused quite a stir. They eventually garnered over $120 million in small donations for political candidates, helping Democrats regain the House and Senate in 2006.

It would seem that the internet was fulfilling its promise as a true platform for the people, but Pariser began to notice something hidden within the interface.

"I had always been excited about the promise of the internet and new technology for making democracy work better," he said. "But in 2008 when I stepped back from being the director of MoveOn, I sort of looked around and saw that it wasn't really turning out that way." After conducting exhaustive research, he penned "The Filter Bubble: What the Internet is Hiding from You," exposing the hidden forces that filter our quest for information.

The book defines the filter bubble as "a unique universe of information for each of us that fundamentally alters the way we encounter ideas and information." These bubbles are works of algorithmic art: functions of coding that enable search engines and web sites to track our clicks via cookies, eventually forming a digitized idea about who we are and what we're looking for. They use that information to present us with relevant results, not to mention well-targeted advertisements.

"Personalization is already much more a part of our daily experience than many of us realize," Pariser said. He first noticed the adverse effects of personalization on Facebook, where his a news feed—a collection of friends' most recent interactions and updates—skewed towards coverage of those whose views were similar to his own. Had his other acquaintances deleted their accounts, or did they simply post less frequently? Neither, as it turned out. Instead, Facebook algorithms had noticed that Pariser clicked on like-minded people more often, and software codes responded by silencing the dissenting voices within his bubble.

If a functioning democracy depends on an exchange of different ideas, hyper-personalization may threaten progress. "The long term result is not just that we're surrounded with ideas that confirm what we already believe, but that we're increasingly unable to see what ideas we're missing," said Pariser. "It gets harder and harder to get outside your point of view when all of these websites are feeding it back to you."

So who benefits most from personalization? For users, an effective filter can certainly be convenient; it helps you find just what you're looking for in what might otherwise be a dizzyingly expansive cyberspace. But it pays off for advertisers too, allowing them to target their marketing with unprecedented accuracy.

The Suit spoke with Martin Feuz of Goldmsith's University of London and Felix Stalder of the Institute for New Culture Technologies in Vienna, two authors of the first peer-reviewed study diagnosing "the mechanisms of personalization within the domain of universal Web search." Their 2011 report, called "Personal Web searching in the age of semantic capitalism," is based on a study in which they created digital identities for three philosophers: Immanuel Kant, Friedrich Nietzsche and Michel Foucault. After surfing the web for months to establish a thoroughly personalized cookie for each of these three identities, the researchers than monitored how frequently—and how well—Google tailored their search results.

In the study's conclusion, they wrote: "This research has indicated that personalization is a far from unambiguous process simply delivering better results to the user. At the moment personalization is both taking place to a surprising extent, but with relatively trivial results." Each of the three philosophers were eventually presented with a unique version of the internet, but those versions did not seem to reflect their supposed interests.

In "The Filter Bubble," Pariser suggests that "the user is content." In other words, end users are not the main beneficiaries of digital filtering technology. Instead we—or rather, aggregations of our web history data, which may or may not reflect actual preferences—are the product being sold.

Stalder explained the concept. "Advertisers clearly benefit from having their ads displayed to a "personalized" and thus more relevant audience… and advertising is Google's core business . Given the lack of clear qualitative improvement of the search results from the point of view of the users, it seems a credible assumption that at least some of the pay-off of personalization accrues elsewhere, that is, on the side of the advertisers."

Here privacy becomes a concern. Although many search engines—Google included—promise never to sell the information they collect, plenty of websites make a profit this way. Pariser notes, for instance, that if you visit a discount travel website, a major airline may catch wind of your interests. Suddenly you'll be seeing their ads for low airfare alongside your favorite blog.

In that scenario, how does this major airline gain access to your data? Pariser points to a new marketing middleman: data collecting companies that auction information off to the highest bidder. He reports that these profitable businesses—the two most prominent now are Acxiom and BlueKai—have accumulated profiles on 96 percent of the American public.

On Acxiom's website, anyone interested in behavioral marketing can search a database to find receptive targets for ads and products. The potential specificity is astounding. It costs $60, for instance, for names and addresses of the 367 women who live in Oregon, are college educated, own a motorcycle and enjoy hunting. Information like this is incredibly useful for the many businesses that make up Acxiom's clients, but it's important for end users and consumers to be aware of their existence.

Transparency is key. "Companies that are collecting data have to make clear that they're doing it," Pariser insisted, "so that you can see when you're getting hyper-tailored results." Citizens need more control over personalization, he argued, before market forces push businesses to use our data in more insidious ways.

Feuz and Stalder agree, proposing a similar solution for Google and other search engines: "They could easily allow their users to swiftly select between a personalized and non-personalized view of their results."

In the end, the onus of resisting an increasingly filtered future rests on the public itself. Pariser warns us to be wary of the "invisible revolution," raising the Orwellian specter of personalization that threatens the privacy of our citizenry. "It is in our collective interest to ensure that the Internet lives up to its potential as a revolutionary connective medium," he said. "This won't happen if we're all sealed off in our own personalized online world."

< Prev		Next >