What prevents us from gaining knowledge? For centuries, our instinctive answer has been: paucity of information. Our search for knowledge was limited by space, time, and resources. The assumption that overcoming this limitation was key could be found everywhere, even in science-fiction. Think of how Star Trek’s Captain Picard learns something new: the omniscient “computer” is asked and authoritatively answers.
We have now arrived at a presumed solution to our knowledge limitation. Over three billion times every day, Google is, effectively, asked a question. It searches among an unimaginable array of information—unhindered by time, space, and resources—and presents to us its answers. And our confidence in the results shows that, as with all newly dominant media technologies, we are absorbing a new epistemology. In other words, we often know what (we think) we know about the world based on the presumed authority and authenticity of a Google search. But is our way of knowing as reliable and objective as we assume it to be?
First, let’s admit that we have muddled what ought to have been self-evident. Increased access to information makes it harder, not easier, to gain knowledge. Why? Think about doing research before the Internet (if you can remember). Your key act of discernment was rejecting what was not relevant. Choosing what to ignore is essential, otherwise your search is both never-ending and inconclusive. With a vast body of information like the Internet, that which does the ignoring and rejecting for you is even more determinative. When your scope is infinite, the filter is king.
The great myth of our Internet age is that it has stripped life of middlemen and diffused power to the many. In fact, not only have the intermediaries merely changed, they are now more concentrated, powerful, and unquestioned than ever. This is not a conspiracy theory, this is the unspoken bargain which most of us make every day. Google and a handful of other Internet giants are the filter-kings, sitting between you and the reality you trust them to bring to you. And it is, therefore, a world they inevitably and unavoidably shape.
How do they do this? The short answer is, of course, algorithms. In the world of programming, an algorithm is simply the code you write to get from a particular input to a desired output. That is a deceptively simple explanation, however. Google has been tinkering with the main algorithm that guides its search engine for many years and the current version—named Hummingbird due to its precision and speed—has at least two-hundred signals, each with around fifty sub-elements, which determine the results. If this is not a middleman, I don’t know what is.
This algorithm is so complex that it eludes comprehension and so commercially valuable that it remains a closely guarded secret. It consists of hundreds of thousands of lines of code and, as Eli Pariser explains in his book The Filter Bubble, “even to its engineers, the workings of the algorithm are somewhat mysterious.” Should the world’s most pervasive and influential filter, serving as a mediator for more than one trillion searches every year, be veiled in complexity and secrecy?
It is also driven by a clear strategy, best summed up by one word: personalization. The link-frequency rubric that directed selection and ranking in Google’s early days still plays a part, but since 2009, personalizing the results to individual users has guided many of the algorithm’s developments (this is true for Facebook, Amazon, and others as well). Based on a wide range of your activity on the Internet, much of which is now tracked and databased, Google tailors the search results you get to the results the algorithm believes you want. Therefore, your search results are not the same ones that your friend will get, and certainly not the same ones your enemy will. They have been filtered specifically for you.
This may sound wonderful. But what does it entail? In addition to your Google searches, the data from your emails is added to the mix. That means if you use Gmail, or a business e-mail provider that routes through Google servers, or even just send e-mails to a Gmail address, everything you write is stored and parsed to refine your online experience. Unnerved yet?
But even this is not enough. Also added to your Google profile is where you go and what you do with an Android-operated phone, every site visited on Chrome, and sites visited on other platforms that use Google Analytics or one of Google’s advertising services—possibly more than half the Internet, according to W3techs.com. We could go on; any service and feature that Google provides is part of an incredible cross-function data-gathering effort that will be, if possible, connected to a specific individual. This is an operation that, in creating a bespoke informational world for each of us, has an insatiable appetite to know everything about everyone.
“Personalization is based on a bargain,” writes Pariser. “In exchange for the service of filtering, you hand large companies an enormous amount of data about your daily life.” It is obvious why personalization is alluring in a self-obsessed, image-conscious age like ours, but why is Google itself so interested? If you ask this, you may have made the mistake of regarding yourself as Google’s customer. You are not. As some observers have pointed out, you are not the customer, you are the product.
Understanding this, Google’s personalization mania and the resultant online bubbles of you-oriented information make sense. The Internet behemoths are profitable and valuable because of advertising. It accounts for an extraordinary $59.5 billion of Google’s total 2014 revenue of $66 billion (and $11.5 billion out of $12.5 billion for Facebook). The reason these figures are so high is because companies are willing to pay more to advertise to pre-selected people already interested in their products. In other words, Google creates a detailed profile of you and then keeps you in a content-bubble determined by that profile in order to sell you to advertisers.
It is this commercial imperative which makes evading online tracking increasingly difficult. For example, Google was caught surreptitiously collecting data from Safari users in 2012. And tracking is no longer limited to the Internet. Google is now working to follow and guide users via smart phones so that it can determine when a product it advertised online is bought offline.
Perhaps, to some, this still seems like a deal worth making; the surveillance is tolerable in exchange for convenience-friendly technology. But the resulting filter bubble that the giant Internet aggregators create for each of us does not just affect shopping. Filtering touches everything that we rely on the Internet to objectively inform us about. It is about everything we think we know.
“The power to include, exclude, and rank,” Frank Pasquale writes in The Black Box Society, “is the power to ensure which public impressions become permanent and which remain fleeting.” When an algorithm is written to ensure that you get what the data leads it to deduce you will like, it is filtering out and downplaying everything it presumes you will not—every disagreeable opinion, viewpoint, fact, and idea. “These new masters of media are more than just conveniences,” Pasquale adds. “Thanks both to their competence and our inertia, they often determine what possibilities reach our awareness at all.”
Consider, for example, the research of behavioral psychologist Robert Epstein. He found that even the ranking, let alone the appearance, of stories about politicians in search results can have a remarkable impact on voting. Years of study concluded that “Google could determine the outcome of upwards of 25 percent of all national elections.” Amazingly, just reading a positive link about a candidate before a negative one made participants more likely to support that candidate.
The billionaires of the Internet era would have us believe that they are objectively unleashing an information-utopia on a world of autonomous learners. And yet the very business model that enriches them helps form a pseudo-reality for each of us that confirms our biases and caters to our thirst for knowledge with methods suited to compulsive shopping. With a mixture of their assumptions and ours, Google’s algorithm—protected from criticism by secrecy and a false aura of objectivity—gives us a filtered reality. It is time to ask the questions of Google, we long ago learned to ask of journalists and politicians: How do you know that? And why are you telling it to me?