Review: Kyle Chayka's Filterworld

It’s Book Week! I’ll be at the Tucson Festival of Books this weekend!

Last fall, around mid-September, someone stuck a bit of driftwood into the sand at a nondescript beach along the Connecticut coast, a stretch of the shoreline that I walk along almost every day. It was a gnarled thing at small scale, but as a whole impressively straight; I’m fairly sure it was machined by human hands, at once point, but had since been battered into an irregular shape, a stick-like shape. By the time it washed up on that beach it had been worn smooth all over its surface, sanded down by what must have been a long journey in the ocean. And some unknown hero had wedged it down into the sand so that it stood straight up. When I first came upon it, that’s all it was - an old weathered piece of wood, standing maybe four feet tall, that some beachgoer had playfully planted like a signpost. I assumed, as I’m sure most people who passed by did, that the stick would fall in short order, blown over by the wind or kicked by a child or perhaps even brought down by the weight of a seagull that might choose to alight on it.

But it didn’t fall. In fact, it stood for weeks. After enough time had passed I ceased to be surprised that it was still standing. I suppose it might have fallen a few times and then been set in place again, but I never saw such a thing, and though I take my morning walks so early I’m usually the only one already up and at the beach, the stick was never lying in the sand when I got there, suggesting it had survived every night. The hole into which it was dug looked sturdy enough, but was nothing special. The stick just wanted to stand. And I was taken by that lonely little act of endurance; there was something oddly comforting about returning every morning at early dawn to find the board still standing there, seemingly indifferent to the surroundings. It’s true that Connecticut beaches are not high traffic, in general, and this one is modest, though lovely. But enough people come by that the fact that no one ever disturbed the stick was inspiring.

Some ran with that inspiration. One day a piece of what looked like floorboard, rounded on the edges and with a jagged hole near the middle, washed up a little further down the beach, and some enterprising fellow took and hung it on the stick. It was remarkable how well they fit together, the stick’s thinner tip poking through the cracked hole in the board, which was close enough to the center that the mass was more-or-less evenly distributed and the board sat roughly perpendicular to the stick. It looked like a lonely T protruding from the sand, perhaps in protest against an authoritarian regime, perhaps waiting for the bus. And the conscious act of a human being to join the two pieces of wood together made it art, public art, free to view, open-sourced. Crucially, hanging the board represented what all true art begins with - taking a chance.

The convergence of the twain pieces of driftwood triggered something in the local community. People began to decorate the stick-and-board crucifix with natural art. The board providing a natural shelf, passersby felt moved to deposit seashells on top like figurines on some dreadfully tacky mantelpiece, first a small orderly line, over time a chaos of dozens. Someone wrapped a line of old washed-up yellowed twine around the top of the stick, above the board, its pleasantly frayed surface bringing a little more in the way of texture to the proceedings. Over time, the stick was adorned with seagull feathers, drizzled with sand, decorated with flowers. Someone took one of the horseshoe crab shells that wash up daily and hung it, just so, off the end of the board. The whole thing was of a piece, a series of natural progressions, and I came to look forward to seeing what was new as part of my morning ritual.

Are you warm, are you real, Mona Lisa?

Then, there was a new iteration, one which gave me pause. We were now deep in Halloween season. Just below the T-junction, skinny against the skinny midsection of the stick, hung a little plastic figurine of a scarecrow-like skeleton figure, complete with a Jack-o-Lantern head, and secured with twist ties to the wood, trussed like a deer waiting for the taxidermy man. It was small, and could not be made out from a distance. When I approached it that first morning I stood staring for an unreasonably long time. (Everyone who lives around here has at this point seen me talking to myself, gesturing wildly, so it would have been no great loss.) On the one hand, it certainly felt like the implied rules had been violated. Though the wood showed clear signs of human manipulation, and the twine had not emerged from the sea full-formed without the influence of man, everything utilized to that point had been natural in a way the pumpkin-head figure was not. Plastic is plastic and though it’s unfortunately common at beaches it does not belong there. The twist ties, too, seemed to break some ephemeral-but-universal compact against securing one’s addition with anything but the materials on hand. And yet I had to give it to whoever had done it - it looked good, right. The little cackling pumpkin face fit the season’s deepening autumn feeling, and the stick already had a taste of the macabre, a spooky essence. The tiny scale of the thing, and the fact that you had to approach closely to see it, felt just right. The little plastic doll was OK by me, and anyway, I had no right to take it down.

Then disaster struck. Three or four mornings after the appearance of the scarecrow, I arrived at the stick to find that it had been ruined by someone with no art in their soul. Apparently inspired by the plastic addition, someone had strapped a two-foot-tall “evil clown” doll to the stick, its size and garish white costume drawing the eye even from many yards away. On the board, this talentless brigand had scrawled (in Sharpie!) the words, “Long Live the Stick 2023!!!” Rarely in my life have I felt so disgusted by philistinism. This was a person who had seen what was happening with the stick, wanted to participate, and when they did so demonstrated that they had never understood what the stick was about at all. The idiotic, Spencer’s Gift’s-style killer clown doll demanded attention for itself, broke the visual flow of the sculptural object, and stood in ugly contrast with the natural materials and muted earth tones of what had come before. The skeleton doll, which I’m guessing gave this vagabond moral permission to do his dirty business, had fit aesthetically and spiritually even if it amounted to a provocation. The clown doll simply ruined the stick, visually, made it into something crass and confused, while the slogan scrawled on top demonstrated a complete lack of understanding when it came to the implied-but-existential rule of this particular artistic project: no one was to call attention to the act of community artistic creation that was occurring there.

I wanted to kick the little fucking doll into the ocean and then dismantle the stick myself, so thoroughly had that last addition ruined it for me. I didn’t feel that I had the right, so I simply walked away, thinking of Oscar Wilde’s dictum that the penalty for bad art should be death. I was satisfied and unsurprised when, two days later, I came back to the beach to find the stick gone, all of it, nothing to suggest that anything had once stood there except the lonely little hole in the sand. I assume someone else was as offended by that last turn as I had been.

Here’s the punchline: I think that experience, the story of the stick, is perfectly symbolic of the awful state of creative and artistic production in the internet era. And I thought of the stick often while I read Kyle Chayka’s entertaining, perceptive, perhaps redundant new book Filterworld.

---

Chayka’s book, subtitled How Algorithms Flattened Culture, is a wide-ranging, well-marinated, occasionally meandering book about how digital life has fundamentally shifted human aesthetic production; it’s an examination of how the algorithms that lie underneath our various platforms and networks have not just transformed online life but in doing so mutated offline life. Where once human being served as the primary curators for our cultural environments, now mathematical abstractions recommend various forms of arts and media for us, dictating fashions, altering style. Amazon pushing a book on us like a fentanyl dealer selling his wares, Netflix serving up yet another project by the delightful Daniel Chang, that same hideous wallet showing up in your Instagram feed again and again, each happens because somewhere some piece of code gave you a once over and decided what kind of person you are. And, in one of the many observations that Chayka makes well that has been made before, this is not merely a digital phenomenon. In the architecture, interior decoration, and photo staging of spaces like Airbnbs and coffeehouses and museums, algorithmic influence has entered the material world, the world of actual things - even provoking changes, increasingly, in our actual bodies. Algorithms don’t have to determine our aesthetic environment, but it’s easier if they do, so we mostly let them. And as ever, capital is watching closely. Where culture goes, capital follows; where capital goes, culture follows.

How do the rudimentary recommendation engines of the 1990s, matching simple subject and title queries to find related items, evolve in time to set trends in architecture? Because people make more of what’s popular and algorithms are the fingers on the scales of what’s popular.

The archetypal example here is Instagram and its influence on the built environment. In the 2010s many of our public spaces started to be molded, physically and literally, by the market advantage of luring patrons looking for highly Instagrammable shots. Businesses that once paid large sums of money to have photographs taken for advertisements that cost even more to run now had access to an army of free labor, taking pictures for free and posting them to a vast network. But to get that free advertising, you had to have a business with a look conducive to Instagram’s communal stylistic values. Your space needed to inspire people to take Instagram photos, photos that fit a very particular set of unspoken but consistent attributes: striking enough to be worthy of capturing, sufficiently minimalist to appear classy, possessing some aspect that seems covetable and exclusive, and (at least for much of the platform’s history) conducive to the square dimensions that the app mandated. Those who were looking to leverage Instagram’s popularity, particularly during its early period of unusual favor with the hip and trendy, had to play by Instagram’s rules, and until 2015 that was the rule of the square.

I think that last part is underdiscussed and very important. Part of what defined the Instagram look was that its images were suitable for a portrait orientation, and portrait orientation is important for an app because turning one’s phone to look at it makes it less likely that you’ll use that app reflexively, obsessively. Form followed function that followed form; what was popular was decided by what people thought looked good but what they thought looked good was dependent on tech industry choices made far far above their heads. This is core to Chayka’s overall analysis - algorithms spring from the specific needs of individual tech companies, those needs are dictated by influences that have nothing to do with aesthetics, and yet algorithms inevitably dictate aesthetic habits; the convenience of a few coders in Silicon Valley becomes more influential on our shared aesthetic values than Anna Wintour. For years visual culture was Instagram culture, to a striking degree, and Instagram photos were square, and so visual culture was dependent on square images, based on decisions Apple made about the form factor of a new product in development in the mid-2000s. This randomness, the way minor practical decisions made with utter apathy to visual culture drive visual culture, haunts Filterworld, and Filterworld.

“Designed and maintained by the engineers of monopolistic tech companies, and running on data that we users continuously provide by logging in each day, the technology is both constructed by us and dominates us, manipulating our perceptions and attention.”

What does any of this have to do with the stick? I think a similar sort of abstract visual reasoning was going on with that community project, and that in an almost literal sense, its development over time was algorithmic. It’s no surprise that the era of algorithms has also been a mimetic era. “In Filterworld, culture has become increasingly iterative,” writes Chayka. You make memes to impress but to impress within the constraints of conformity; what you’ve seen is what you get. A great deal of contemporary human creation, if you can call it that, has been in meme form. Someone comes up with a juxtaposition of image and text, people iterate on it, literal templates are made to facilitate easier reproduction, and the immense distributed artistic collective of the internet produces copies of copies of copies. Whatever ingenuity and surprise was found in the original meme is quickly drained by this process. And not only do memes go stale, in my experience and my opinion they go stale quicker and quicker; the process of mimetic exhaustion accelerates. Mine does, anyway.

Memes and algorithms are deeply intertwined. In the most basic sense, social media algorithms almost always reward that which has previously been popular, has previously received attention; memes are simply the repetition of image formats that have proven recently popular. Algorithms are sets of quantitative rules that dictate computational behavior, while memes are sets of implicit rules that dictate social-creative behavior. Both have inputs and outputs and expected limits on what can been altered. Each orders expression. And so with the stick. As with a meme, the gradual construction of the art piece that was the stick was an emergent process that stemmed from social rules. Every person who added a shell or twig was playing within a form that had been set in motion by whoever had thrust the stick into the sand, and the iteratively-developed “rules” of the stick were similar to the types of alterations to a meme format that are socially accepted. (An expression of a meme that does not conform sufficiently to the communal understanding of that meme’s rules is likely to be rejected, not seen as an example of that meme at all.) If an algorithm is in its most basic form a stepwise process, a series of “if-then” statements, then various aesthetic genres and media and traditions inevitably become algorithmic over time. The great artistic challenge is to violate expectations and defy tradition while in some essential way working within the formal constraints that make art possible. This is very difficult, and thus being a great artist is very difficult.

I was offended by the last evolution of the stick, in an incredibly low-stakes way, because I felt that it was ugly, but also because I felt that its ugliness stemmed fundamentally from a failure to understand the implicit rules at play. And yet I recognize the confusion and hypocrisy of the situation, first because I am someone who finds memes terribly boring because of their basic repetitive nature and second because I have always been an admirer of the avant garde, those who violate expectations most flagrantly. I want people to get more creative, to break the rules. But as that hideous clown reminded me, to break the rules effectively, you must do so well. And some of the questions of the algorithmic cultural age include when and how we can break the rules imposed socially and in software, and which ones, and also whether we remember how to break them well, after so many years of mimetic art, relentless nostalgia, and copycat culture.

Filterworld made me think of the stick because the book is at heart a deep examination of the forces that shape artistic and cultural production, and in particular how the rule-bound processes we live under subtly define what we can create. Chayka’s book is wide-ranging and subtle, informed by a lot of world travel and interviews with interesting people, and I enjoyed it a great deal. I also struggled to avoid the temptation to make the subtitle of this review “The Best New Book of 2018”; that’s a testament to the trickiness of writing recent history and to the degradations of the glacial speed of publishing.

---

Chayka begins Filterworld with the story of the Mechanical Turk, which (as you’ve no doubt read about before) was a fake robot, a false automaton, a fugazi android. In the 18th century the German inventor Wolfgang von Kempelen, looking to impress the nobility, “invented” an automated chess playing machine, which was able to beat most challengers. This bit of deception depended on a small man hiding inside the machinery and moving the pieces from below. (I can see the job listing now, looking for an unusually small man with an ELO score over 2000.) The ruse was so impressive that von Kempelen ended up taking it on a European tour; the hoax lived on for fifty years after von his death. I’ve sometimes wondered whether this was just a crazy scheme that got away from him, like a sitcom character. Chayka likens our modern algorithmic systems to the Mechanical Turk, in that they too are ultimately an expression of human choices pretending to be a machine’s decisions. By empowering the algorithm, these corporations hand over power to a seemingly indifferent outside force; hey, we didn’t censor you, we didn’t harvest your data, we didn’t fill your feed with stuff you never asked to see, the algorithm did. It’s very convenient!

A little later, Chayka mentions Project Cybersyn, a bit of Allende-era stagecraft put together by the Chilean government. They designed an elaborate facsimile of what’s immediately identifiable as a 21st-century command center, tons of data at the fingertips of leaders whenever they wanted it. Of course, at the time the technology needed to pull this off did not exist, not in Chile or anywhere else. But just about everyone admits that it looked really cool. Here too you have the “do not look at the man behind the curtain” reality of much of our modern age; deception is woven into technology because human decisions can be hidden in the immense complexity of digital systems, obscuring human manipulation. In mentioning these stories Chayka invokes the artifice that’s so deeply woven into our online lives. I think immediately of the Facebook moderators who work long hours in distant countries for little pay, making human decisions about what to keep up and what to remove, which most people assume is done algorithmically, a little army of Mechanical Turks in the third world acting as a shield for those in the first.

As the title suggests, Chayka is largely uninterested in algorithmic influence on the economy or government. He’s instead interested in how algorithms mold arts, media, and culture. Necessarily, this means that he devotes a lot of time to the overwhelming tide of recommendations we live in, recommendations that tell us what show to watch next, what book to read next, what useless junk to buy next, even which pornography to consume next. Core to Chayka’s point is the obvious but essential fact that these algorithms can do little else but recommend to us that which other people have already clicked on, suggesting an implicit (and deeply ideological) assumption that we always want more of what we’ve gotten before. If it’s true that our culture can create nothing new, maybe it’s because everyone is constantly fed a steady diet of stuff that’s like the other stuff that was like the stuff that came before it. “Under algorithmic feeds, the popular becomes more popular, and the obscure becomes even less visible,” Chayka writes. The fact that code is forever telling me that I’ll like something because I liked something else that’s superficially similar is deeply insulting. I don’t come close to fully understanding my taste. Amazon certainly doesn’t. But they’ll keep up with this presumptuous business, based on those dubious connections, and there will always be more because there’s just so much of everything. Filterworld lingers over the point: this makes the recommenders - the algorithms - immensely important to anyone who creates.

One of the more bitter elements of contemporary life, for many people of an artistic temperament, is that the means of creation have been democratized (with affordable and accessible video and audio recording and editing equipment, for example) while the platforms where one might become successful are so choked with entrants because of that accessibility that it’s next to impossible to get noticed. Trying to become a prominent YouTuber depends on massive amounts of luck as well as skill, and the sheer, impossible mass of the number of people trying to earn not stardom but just a stable audience is incredibly discouraging.

I get asked for advice from writers committed to doing this professionally, all the time. They tell me that they think their stuff is good and they feel certain that they could get an audience if they could just get a break, and ask me, could I tell them how? I generally always decline to look at their stuff, because there’s no percentage in it for either of us - I’ll have to tell them that they’re not good if they aren’t, which I don’t want to have to do, and if they’re good it makes no difference since I have absolutely no juice in this industry with which to help them. And I can’t tell them how to make it because I have no idea. Being good, in this industry, is neither necessary nor sufficient. I didn’t get in that early, but it was early enough, and I connected with the right people at the right time, and I caught some lucky breaks and bounces, and a couple times editors who might easily have told me to buzz off helped me out, and also, yes, I’m very good at this. Do I think that, if I were to start today with the exact same level of skill and ability and knowledge and education and work ethic that I have now, I could confidently say I would eventually rise to this already modest level of attention? Not at all. Maybe I would catch the right breaks. But look at this newsletter platform, right now, stuffed with talented people who can’t get a foothold. My own secret superpower is that I have modest ambitions, and I think anyone else who’s trying to make it in a creative field should labor to hold them too.

There’s a sense of gentle offense in this book, which is never polemical but amounts to an indictment all the same. And I think it stems from this feeling that we’re all lost in the mire of each other’s ambitions, our creative output twisted by what depersonalized and accountable systems reward, and still buffeted by chance. Chayka is very clear that the older eras of human gatekeepers were deeply imperfect, filled with all of the biases of petty human bigotry, distorted by favoritism and corruption, fickle, and often just tasteless and wrong. But at least there were actual human beings at the other end of those decisions, not data flowing through servers making consequential decisions for all of us, working ultimately only to the benefit of a small handful of zaibatsus.

There are alternatives. The last couple of chapters discuss various ways that the pernicious elements of Filterworld could be ameliorated. One is about regulating platforms. I’m in favor, but pessimistic; these corporations are too well-moneyed and too powerful, and it’s very difficult to implement meaningful oversight on processes that occur many thousands of times a second. More convincingly, Chayka discusses weaning ourselves off our feeds a bit, describing his own fitfully-effective efforts to do so himself. It can feel hard, and many elements of modern professional and social life want to push us into the recommendations feeds; good luck with online dating, job hunting, or even simply using email without ever finding yourself with a recommendation you didn’t ask for. But you can take steps to pare things down. Chayka has made some changes to limit his time in Filterworld, though he’s self-deprecating about how hard it is for him and how limited his success has been. I’m still floating around in this online hellhole myself, obviously. But it’s also true that I no longer have any true social media, no Facebook or Instagram or Twitter, and while Letterboxd and Goodreads may seem to be online recommendation engines of just the sort that can be troubling, ultimately the actual reviews are written by humans. And, anyway, I don’t use those services to figure out what I’ll read or watch next, I use them because I didn’t already have enough places to hear myself talk. Anyway: Filterworld can be minimized in your life, though only by truly disconnecting from the online experience can it be escaped.

Incidentally - well, OK, the opposite of incidentally - Chayka stresses that he thinks that the influence of technology has been felt in culture forever, which is plainly true. He also says that this influence is neither good nor bad but neutral, and here I confess I don’t really believe him. It’s not that I don’t believe that it is neutral, myself. I mean that I doubt that he thinks it’s neutral, given the rest of the book he’s written. But I could be wrong.

---

I have tried to think of exactly how to put this ever since I got deep into the book, and that kind of effort to find my words is almost always an indication that I’m being either too clever or not clever enough. But here goes: the sense of mimetic exhaustion that stems from encountering the same forms over and over again is present in the text of Filterworld itself. Chayka is always engaging. The trouble is that the subject matter frequently feels very dated. Must we, again, have the conversation about how every coffeeshop started looking the same sometime in the early 2010s? I guess we must, given the book’s purview. And I guess we must talk about menu items at restaurants that exist only to be Instagrammed, and entire bars painted pink for the same purpose, and so on. Still…. Of all the places to go to see how the world has been transformed by digital culture, do we really need to go to Iceland, the rapid transformation of which has been covered again and again? All of this fits sufficiently well into the book’s purview that, were it not included, I would probably note the absence as a detriment. But getting through it sometimes feels like a slog, to go over the same material once again, like going on the Universal Studios tour for the fifth time. Chayka refers to an article he wrote about this stuff in 2016, which feels like a bit of an own goal. The Instagram section indeed feels like it’s ripped from the pages of eight year ago.

I like all of these pieces braided into a single cord, and there are nuggets here I hadn’t previously considered as well as a kind of panoramic perspective that’s rarely brought to bear on life inside our smartphones. But I have read a lot of this before; indeed, I have read a lot of it before from Chayka. And if you want my most basic “should I read this book?” wisdom, I would say that my answer depends a great deal on your patience for treading this old ground. For me, the overall package was great. But I can see a less patient reader leafing ahead to find topics that have never been discussed in an episode of This American Life. In three successive pages, we get “We are all influencers,” “People are platforms, too,” and “The primary incentive is to make the [follower] numbers go up,” the last a part of a treatise on how social media reduces people to numbers. These things are not untrue! The repetition of 2010s conventional wisdom did make me go oof, though.

It is of course Chayka’s right to write a work of recent history; it’s just a tricky thing to do. News is inherently interesting. History of the sufficiently-distant past is often inherently interesting. Recent history, not of secrets and backroom dealings but the recent public history of widely-accessible aspects of our daily lives, experienced by the vast majority of Chayka’s potential audience, must be made interesting, and this is hard work. It’s made harder by the unhurried pace of the publishing world. And I can sympathize. My recent book suffered from the ceaseless movement of present into past and, in a strange way, from the algorithmic nature of contemporary politics.

How the Elites Ate the Social Justice Movement sold well enough to not be an embarrassment, but didn’t come close to earning out my advance, making me a two-time loser on that front. (The Cult of Smart keeps creeping up, though, and has achieved something like sleeper status, so maybe there’s hope. You should buy both, as well as a copy of each for a friend.) The irony is that Elites was my sellout book. I believed and believe everything that it says, I think its message is important for the contemporary left, and I think it describes very real and persistent issues that hamstring the fight for social justice. But it’s also true that I wrote it because it was the book that everyone expected me to write and wanted me to write, and because it was the one that would get me a big advance. It never popped in the way Simon & Schuster thought it would, though.

Why? To connect to the theme of the exploitation of algorithms, I thought that sales would proceed naturally from the rage it would induce in our media class. I understand that this is cynical but it’s a cynical business. The marketing logic, at least on my end, was always predicated on the book receiving negative press on Twitter and from there generating publicity. The book was, as a political journalist I trust deeply told me, relentlessly sympathetic to the protesters, but it was still fundamentally an act of criticism and I knew that Twitter was not the kind of place where nuance was common. I got all the formal attention I wanted, in newspapers and podcasts and similar. What I needed was organic people power. Everything about the book’s marketing seemed likely to infuriate the media professionals who held such outsized influence on Twitter, and they are, as a class, people who are always willing to help publicize that which they hate.

When Pamela Paul wrote a column about the book, I knew that what I needed had been accomplished; “Pamela Paul writes approvingly about Freddie deBoer’s book criticizing BlackLivesMatter protests in The New York Times” is the Twitter nuclear codes. I struggled to imagine a more inflammatory string of words in the English language than that, for Media Twitter. It was a thing of beauty, really.

And yet the column went almost entirely unnoticed in that part of the world, though I’m sure in general that bit of publicity moved some units for me. Book sales were lackluster. Why? Because the world I had been writing for changed, in a way that speaks to both literal algorithms and the more metaphorical algorithm that Chayka sometimes refers to. For one thing, Elon Musk bought Twitter and likened it in his image. He stripped the legacy bluechecks of their status symbols. (It pains me to agree with Musk and his cronies in this way, but it is absolutely, positively the case that getting a blue check was a mark of status in digital media, it simply was, stop lying to yourself.) He gave those checks to those who would pay for them, which ended up being people who aligned themselves with Musk ideologically, and that was a group that tended towards the vociferously conservative where Media Twitter had been almost monolithically liberal in both the good and bad ways. He also throttled engagement for tweets that linked to all manner of publications, including, at least for awhile, the Times. Meanwhile, the people I had expected to get big mad about Paul’s column, that Media Twitter bloc that had been so recently such a hegemonic force, had suddenly been broken up by these tectonic changes, their engagement throttled, their reach limited, some left to take refuge in one of Twitter’s many pale imitators. There was no longer a critical mass to get conveniently mad at me.

A book I had pitched according to rules that made sense in the early days of 2022 was suddenly orphaned by the rules of fall of 2023. Whether it would have sold or not, I don’t know, but the book’s inability to generate useful online outrage in light of algorithmic changes was an outcome straight out of Filterworld.

And so I definitely sympathize with Chayka’s plight. Publishing moves glacially slowly. (I actually wrote my first book, published in 2020, as a high school sophomore in 1997.) Books designed to describe particular moments, as Filterworld is at least in part, perpetually feel just a step behind. There’s more than enough in Chayka’s book that’s timeless or forward-looking for the journey to be worthwhile. Did I sometimes get impatient while reading about social-technical dynamics that have been examined for a decade, including by Chayka? Yeah. For the record, a lot of subjects fall into this weird valley where they seem both too recent and too dated to be covered in nonfiction books. But feeling of the moment certainly isn’t the only virtue a book can pursue.

---

At the heart of the thing, luckily, lies the kind of perpetual curiosity that’s essential for writing a good work of nonfiction. If you don’t have the capacity for being repetitively interested, taken with new aspects of your subject again and again, I don’t think you’re going to make it. If Chayka’s subjects are well-worn and his observations sometimes obvious, his book is consistently powered by his cheerful commitment to holding his subject out at arm’s length and twisting it around - clockwise, counterclockwise, into different light. The curiosity in the book, for me, was infectious. Perhaps it’s best if you think of Filterworld as Chayka taking a stately walk through a mall of the recent aesthetic past with you, popping into stores occasionally to pull an item of the rack and talk to you about it. Surprisingly, it’s a book that strolls, surprising because it’s about (among other things) the ceaseless acceleration of culture.

I often felt, while reading, that Filterworld wants at heart to be the product of a little museum press, more explicitly academic and less concerned with being commercial, freed from the addiction to newsworthiness and commercial relevance that dog high-profile publishing. (But I also understand that writers need advance money, so.) Chayka’s basic interest in algorithms doesn’t lie in their nature as forms of math or objects of computer science. Still, I could have maybe used some more about who exactly makes these things and how, not in any technical terms but rather just a sense of who the actual nerds in the back room are, dictating the orderly sorting that algorithms busily do thousands of times a second. But I recognize that that’s all incidental chemistry. Chayka is, instead, keenly interested in the aesthetics of his moment and the platforms that have so witlessly shaped it. This is fundamentally a book about taste. What I think is most useful in Filterworld is Chayka’s understanding of a concept that he does not explicitly mention: path dependency. Aesthetics and fashion are products of choices born of choices born of choices, a flow of causality that nobody planned and that you could never perfectly predict.

When I was in high school, an instantly-identifiable kind of green cargo pants became popular. The story (or myth) goes that this followed a peculiar pattern in fashion, the way clothes that are embraced because they are cheap become popular and then become expensive. Army surplus stores had many pairs of these green cargo pants for sale, that cost next to nothing, and this attracted cash-strapped buyers looking for a good deal; the look caught on, designers began to incorporate it, and the next thing you know green army cargo pants could be bought from high-end fashion houses for $800 a pair. Chayka tells the story of how the bob haircut came into fashion because of one French woman’s dogged efforts to get rid of an old boyfriend who couldn’t take the hint. These may only be algorithmic processes in the most abstracted sense, but my suspicion is that it’s precisely that kind of process that most animates Chayka in this project.

Path dependency has always directed taste. The trouble is that now a tiny handful of massive tech firms dictate those paths, in a way that has indeed flattened culture. An algorithm says if, then; the user might decide the “if” but the platform always decides the “then.” The input might be human, but the output is never firmly under a human’s control. What Chayka’s done is to spend a lot of time wondering and worrying about that condition, never too urgently, not in a particularly novel way, but rather with an obvious conviction that such wondering and worrying about the world around us is never time badly spent. And the book proves him right.