Filterworld: How Algorithms Flattened Culture by Kyle Chayka. Doubleday, 304 pp., $28.
The algorithm has our attention. It befuddles and bewilders, surprises and preoccupies, sucking us in like a vacuum, and spitting us out like a blowhole. Dayna Tortorici understood and articulated this better than many of us ever could in her hugely popular essay “My Instagram,” which captured and perpetuated our frenetic zeal for the algorithm and our discussion of it. Running in issue 36 of n+1 four years ago, the piece is the fourteenth most-read essay in the first twenty years of the publication’s history. Through the story of how she slowly but methodically became addicted to Instagram, Tortorici depicted a pivotal moment in time: roughly 2013 to 2015, when Instagram was still in its near-Beta phase, trying to mimic through its icon and filters the design and effects of a Lomography camera. We were all just beginning to realize the addictive power the algorithm might one day wield. Tortorici’s recounting of her discovery of Instagram’s world of superhuman fitfluencers pulled readers into her palm-sized portal, the rapid zigzag of her prose reflecting what it feels like to scroll ad infinitum.
Tortorici’s essay describes many of the same phenomena Kyle Chayka addresses in his latest internet history, Filterworld: the weird reverse mimesis carried out by our phones, wherein bars and cafés in certain neighborhoods feel as though they were designed to be photographed; the way the internet has become indispensable to our consumption habits; and the Big Tech puppeteers behind the dancing dolls, the astonishing valuation of their companies, and the government’s slow but eventual intervention into the industry’s data-sharing practices.
Where Tortorici’s essay differs from Chayka’s book is not only in its onomatopoeic qualities—and its ability to capture a wider range of material, earlier and in far fewer words—but also in its understanding that the algorithms are more our mirrors than our captors, that our interests motivate us toward the algorithm in order to discover more, not settle for less. “Instagram grows on subjectivity like a fungus whose shape and color varies from person to person,” Tortorici writes. “To describe what it feels like to live with it is not to describe how it works.”
Let’s start with an example. Sometimes when we’re wasting time on the couch, I ask my partner if I can look at his algorithm, by which I mean his Instagram. My “Discover” page usually looks a little different from his: disabled animals, dogs in hats, too-expensive shoes, pretty girls in Berlin styling their “outfits of the day,” midcentury sconces and credenzas. My partner’s is mostly nonsensical memes tracked with songs that blow the speakers out; Tony P, the “25-year-old bachelor living in Washington, DC”; a guy eating a bowl of hot ramen while riding his bike; someone playing John Coltrane’s “Giant Steps” on a plastic recorder. It’s fun—and funny—to discuss our algorithms because they’re a little bit personal, private even. We all know that they are a reflection of our personalities in some way, or at the very least, a reflection of how we like to delight in mindless entertainment. (As Tortorici puts it, what crosses your feed is nearly psychoanalytic, “like blurting out a dream.”)
This kind of machine learning can be helpful when you’re on the internet to look for something specific. I know that it is seamlessly, furtively assisting me when I go hunting for something tangible on Facebook Marketplace (a piece of furniture) or something slightly less tangible on Instagram (evidence that someone I haven’t spoken to in years still exists, is married, and has had a baby). But most often when I open my phone, I am not looking for anything in particular. Usually, I am simply seeking an escape. The act of unlocking my screen and tapping the app icons is reflexive, muscle memory—something I do without even noticing I’m doing it. And then I’m there, and I find a piece of news or a purchasable item that interests me, so I stick around and perhaps enter my credit card number in one click. I continue clicking until I find something that produces the same level of excitement. I recall how Anna Weiner wrote in Uncanny Valley (2020) that it’s no wonder the tech world calls us all “users.”
In Filterworld, Chayka tries to explain how we got here, but not in a way likely to satisfy anyone well-versed in the ways of the internet. While algorithms and our desires are a classic case of chicken-or-the-egg, Chayka leans into the idea that algorithms come first, while what we discover, choose to like, and buy as a result, mostly comes second. I might be getting served well-shot videos of a midcentury modern credenza because I have indeed been searching for one, but according to Chayka, the algorithm is probably why I want one in the first place. In what he deems the “Filterworld”—or internet time and space, the setting upon which some opaque calculus has dictated our every move—the algorithm is so all-encompassing that it shapes the very basis of our tastes themselves and, in the process, squashes the curiosity and ingenuity that our culture once possessed—and that the early internet, the fun internet, once helped to promote.
In suggesting that individual tastes have gone out the window, leaving us with a gray, “flat” culture, Chayka gives very little credit to his reader.
Chayka begins Filterworld by describing what he has identified as the precursor to the algorithm: “The Mechanical Turk,” which was thought—mistakenly—to be the first working robot to imitate and anticipate human behavior. He then moves into explanations of how algorithms work, the influence Big Tech and monopolistic growth have had on the internet and our lives, and how the algorithms created by those companies have impacted our taste and subsequent consumption. The book’s chapters move not so much chronologically as from “how we got here” to “how we can get out.” In many ways, it functions as a history of Web 2.0, combining research and reporting with moments of personal reflection, especially as they relate to Chayka’s own internet addiction, the algorithm cleanse he underwent for his research, and his early days exploring the web as a teen, back when “social media was … a niche itself.”
Chayka calls for the return of that kind of internet: one that is more decentralized and less homogenizing. If the early internet, the one of his childhood, was democratic and free-range, today’s iteration resembles more of a fascist structure. The all-knowing algorithm of today, Chayka says, “creates templates of how things are supposed to be, always formed by inherent biases—a bracketing of reality that is then fulfilled by users creating content that fits the mold.”
Most readers who, like Chayka, have grown up with the internet, will consume this information as true, if perhaps obvious. Who hasn’t felt that the face mask that keeps popping up on our feeds must be the result of a multi-level marketing scheme? That phalanx of blonde women in workout classes all wearing the same Alo Yoga set popularized by some TikTok influencer certainly has fascistic undertones. But where Filterworld falters is in its insistence that the algorithm has “removed the burden of choice and the more intentional process of selection that had to happen in earlier eras of digital culture.” The algorithm has certainly made finding things that we already wanted easier, but has it not also created an entirely new realm of discovery? In suggesting that individual tastes have gone out the window, leaving us with a gray, “flat” culture, Chayka gives very little credit to his reader. Viewed through Chayka’s lens, she has little inherent agency, ability to form personal values, or offline community that nourishes her interests and desires.
This analysis could be explained by the fact that Chayka has dedicated much of his career to spotting trends and attempting to diagnose their origins. He does so regularly in his New Yorker column, Infinite Scroll, where some of Filterworld’s ideas first appeared. (See “‘Emily in Paris’ and the Rise of Ambient TV.”) Before the New Yorker, he started Dirt, a Web3 newsletter funded by NFT sales with the writer-cum-businesswoman Daisy Alioto, who has described what they created together as “the future of media.” Covering everything from streaming to sock recommendations to Sofia Coppola, Dirt describes itself as “a daily(ish) newsletter about digital pop culture.” In one of its early dispatches, from 2021, Chayka wrote about how “Uniqlo is Sally Rooneycore.” “You can read a Sally Rooney novel anywhere. You can wear Uniqlo anywhere,” he quipped. The argument was thin, but the touch points—gray-and-white normcore, a literary fiction author turned mainstream—hit the zeitgeist in all its right places.
It is this expression of the zeitgeist, I think, that makes people interested in Chayka’s work. His first book did that in an interesting way by applying his background—art history—to a wider cultural phenomenon—minimalism. In that book, The Longing for Less (2020), Chayka argued that pioneering Minimalist artists like Agnes Martin, Donald Judd, and John Cage were the precursors to a broader cultural interest in minimalism: sparsely decorated cafés, Ikea chairs, Marie Kondo, downsizing as a balm in a world obsessed with consumption. In Filterworld, Chayka applies a similar framework to his explanation of the way algorithmic recommendations have shaped our interests. Here, he invokes the museum curator: that person with a trained palate who is tasked with organizing cultural products and bringing to our attention things worthy of it. Now, “machines are trying to become curators,” he says. And since “monopolistic growth is more important to these entities than the quality of user experience,” who are we to trust the computer’s taste over a human’s?
It’s a valid question, but the invocation of fine art and the museum is much less convincing in this book, in part because the prior example—Minimalism—was a product of a particular time and place, but also because the precursor to the algorithm—the curator—hasn’t disappeared as much as Chayka maintains it has. Certainly, for some people, cultivating taste today is “an uphill battle compared to passively consuming whatever content feeds deliver to you.” But the discerning (and perhaps less lonely) internet user, who already knows that what Mark Zuckerberg feeds them is designed to steal their attention and money, is more likely to be swayed by friends, curators, mentors, Subreddits, group chats, and other communities on- and offline. It’s still the people, not the robots, who have the greatest influence on our tastes.
For Chayka, the algorithm is so all-encompassing that it shapes the very basis of our tastes themselves and, in the process, squashes the curiosity and ingenuity that our culture once possessed.
I will be the first to admit that sometimes it is nice to have a tool that can whittle down a decision in a field of endless choice. If you’re having friends over and want a reliable meal, you might turn to Alison Roman’s website or a Molly Baz video for a recommendation about what to cook for four. If you need to buy people gifts for the holidays, you might turn to the Strategist, where a spate of stylish editors and celebrities recognized for their good taste in one area or another have already tested products and can suggest them with explanations and personal-use cases for each. For reading recommendations, I always turn to my favorite writers on Twitter or the magazines I subscribe to, and for help finding a particular article of clothing, I turn to one of three trusted newsletters written and curated by people with good style who have done the work for me. Some of these celebrity chefs and writers have been brought to my attention by “the algorithm,” based on demographics and search histories that reflect my preferences or those of people very similar to me. But the best recommendations don’t come from lists or parasocial relationships at all, but from real connections: from the friend who is practical or has particularly good taste in shoes, from someone I know who does a lot of research before buying skin care products, from my mother, who knows what kind of curtains I would want, due to the size of my apartment and my sleep neuroses. If personalization is key, who better than someone you actually know to provide you with tailored suggestions? Besides, there are too many products and too many people peddling them online to actually decipher what is worth it and what is not. It is no surprise to me that some people get so overwhelmed they ditch social media completely.
For Chayka, the internet is no longer a place for exciting discovery, but one that has been hollowed out and narrowed, to the extent that no individual is much of an individual anymore. One of the sources he draws on to make this point is a woman named Valerie Peter, whose comments he found in the fashion critic Rachel Tashjian’s newsletter, Opulent Tips. (Tashjian’s newsletter, Chayka fails to highlight, is much like those mentioned above, offering product and reading recommendations from a real person whose lifework is fashion.) Peter complained that the algorithm had made it harder for her to figure out her personal style, and in a subsequent interview, she told Chayka that she had felt completely hypnotized by her apps. She had impulse-bought a pair of leg warmers despite having never liked their look and began feeling intimidated by the astrological messages passing over her screen, even though any interest she’d had in astrology had been fleeting at best.
Chayka uses Peter because she fits his archetype of someone whose identity and interests have been dictated by the algorithm, but reading about her, one can’t help but think that people this persuadable, this easily convinced to buy an item of clothing they hate, are probably no stronger in their sense of self than the teen in Mean Girls who buys army pants and flip-flops because Cady Heron wore army pants and flip-flops. Advertising and corresponding trends have always gaslit us: They’re what persuaded women to buy cigarettes in the 1960s and everyone to get on Prozac in the 1990s. The difference now is that the advertising is attached to our bodies.
I don’t blame Valerie Peter. There are myriad forces far beyond our control pressing down on us every day. Companies hawking us stuff we don’t need via paid influencers have fostered a competitive consumption culture that creates an unyielding number of material desires. (In one of the latest videos from my Berlin fashion influencer, she explains, without a hint of irony, “what the cool girls will be wearing this spring,” a horrible remark for someone impressionable like Peter to consume.) It’s not our fault. But it is up to us to resist those forces, and the best way to do that is to acknowledge their influence and not give them, as Chayka does, more credit than they’re due. And I’d like to think that most people who use algorithmic tools are smart enough to have already done so.