Elis Pariser coined the phrase "the filter bubble"
You cannot aggregate taste. But a flood of rating systems and collectivized percentage values, which guide us toward TV shows on Netflix or songs on iTunes, would argue otherwise. Meanwhile, as newspapers and magazines continue to shutter, we may be in the midst of discarding the few really professional critics we have left. What might this mean for the way we think about art and culture?
Surprisingly, Matt Atchity, editor in chief over at the film review aggregator Rotten Tomatoes, agrees that we should worry about this. I told him I don’t love the idea of aggregating critical opinion, saying, “In some ways it’s anathema to the whole point of criticism, since it strips the critic of a subjective voice.” And Atchity told me, “My worry about that is the one thing that keeps me up at night.” I asked him how he thought of his own role in critical debates, and he told me his job is to amass the best opinions in the country for his millions of readers.
“Sometimes I feel like I’m the town crier,” he told me. “I feel like I’m a herald.”
But doesn’t such aggregation means its an algorithm that gets to be the actual herald? Atchity, in the long-run, may turn out to be more accurately described as that algorithm’s right-hand-man. Such shifts in control lead us into something that Internet activist Eli Pariser has coined “the filter bubble.”
Since 2009, Google has been anticipating the search results that you’d personally find most interesting and has been promoting those results each time you search, exposing you to a narrower and narrower vision of the universe. In 2013, Google announced that Google Maps would do the same, making it easier to find things Google thinks you’d like and harder to find things you haven’t encountered before. Facebook follows suit, presenting a curated view of your “friends’ ” activities in your feed. Eventually, the information you’re dealing with absolutely feels more personalized; it confirms your beliefs, your biases, your experiences. And it does this to the detriment of your personal evolution. Personalization—the glorification of your own taste, your own opinion—can be deadly to real learning.
Shall we engineer instead a kind of critical vacuum, an artificial absence of voices, in which a few educated, comprehensible, and highly subjective opinions can again prosper?
Perhaps we’ll get more of a critic vacuum from companies like Songza, the music-streaming Web site that delivers playlists curated by experts (and occasionally celebrities, from Justin Bieber to New York City’s former mayor Mike Bloomberg). Songza is founded on a simple enough premise: If there are twenty-four million songs on the shelf, people become baffled by the panoply of content and fall back on the few songs they already know; access to everything encourages exploration of nothing. Songza’s job is to ask you what you’re in the mood for (taking a sunny stroll? preparing for bedtime?) and then introduce you to music you didn’t know you wanted for the occasion. It’s an approach that’s working. On any given day, seventy million minutes of activity are logged on Songza. I spoke with the company’s cofounder Elias Roman, a wunderkind from Queens who’s found his way onto the Forbes “30 Under 30” list.
Roman’s ideas about the future of musical decisions will relieve those of us who despair at bossy algorithms. “Some things are easy to crowdsource,” he told me, “but when you’re interested in constructing a playlist, a coherent whole, it’s more than just aggregating a bunch of binaries. I’m saying that there is a value to tastemaking.”
Tastemaking? The very term sounded antique, wonderfully elitist, coming from the founder of a digital start‑up. “We have a desire here to be tastemakers,” he continued. “While our algorithms will sometimes offer music that a user has chosen in the past, we have a mandate that the site always brings forward songs you don’t know you want yet. There’s always going to be both comfort food and something surprising.”
Roman’s insistence on tastemaking flies in the face of most content providers, who seek only to gratify the known desires of users—only to increase clicks for their data mining purposes. We need sites like Songza to continue insisting on “surprise” content if we’re ever to escape the filter bubble. Praising and valuing those rare expert opinions may still be the best way to expose ourselves to the new, the adventurous, the truly revelatory.
Michael Harris is the author of The End of Absence: Reclaiming What We've Lost in a World of Constant Connection.