In Filterworld, Kyle Chayka’s exploration of the profound cultural transformations and distortions wrought by algorithmic recommendation, we visit three bookstores. The first is an Amazon storefront in a shopping precinct in Georgetown, D.C. It’s a bizarre retail experience:  

All the book covers were facing outward, side by side on the shelf like so many digital icons. And each book had a digital label below it showing its respective rank on the Amazon site, with numbers determined by user review, volume of sales, and even how many pages buyers of the book got through, a metric measured digitally by Amazon’s Kindle e-readers. 

Books are arranged into categories prompted by the Amazon website, such as ‘top sellers’, ‘most popular’ and ‘4.5 stars and above’. There are even printed labels nestled between books that read ‘if you like ← you’ll love →’. Chayka is discomfited by this literalisation of the digital environment, as if the technology of algorithmic recommendation had itself come to life. Forget the romance of the bookstore as a refuge, or as a site of discovery, this is a place that a consumer visits in order to be sold books. Chayka, at least, isn’t buying. He’s both overwhelmed and unimpressed: ‘the bookstore selection driven by the average of all of Amazon’s data was curiously homogenous and ultimately boring’. His unease is paradigmatic of the experience of consumers in what his book dubs ‘Filterworld’: ‘surrounded by content, but inspired by none of it.’  

A trip to the Soho outpost of McNally Jackson provides a wholesome contrast, an opportunity to pit algorithmic logic against the work undertaken by human tastemakers. Tables stacked with books at the heart of the shop soothe Chayka and reaffirm the virtue of human judgment: ‘many of the books stacked on display were new releases, but they were also a selection of the books the store’s staff decided were worth a closer look, a collective act of curation… The arrangements always seemed to say: Just trust us.’ Again, this contrast becomes a paradigm for understanding the way consumer choices are formed: ‘through our algorithmic feeds, we only get the Amazon retail experience, not the McNally curatorial eye.’  Chayka wants his readers to cultivate their own curatorial eye, to reclaim an agency he argues is imperilled by algorithmic recommendation, so that they can make cultural choices that enrich their lives. 

Like Chayka, I would prefer to browse the staff recommendations at an independent bookstore than to spend any time at all in the Amazon bricks and mortar shop. When I do visit a bookshop, I am also guided by my own inclinations and affinities, which have been shaped not just by a lifetime of reading, but by the recommendations that come my way when I read criticism and talk to friends. I’m happy to be surprised by new books I didn’t know were on the market but often I’m frustrated by these curated displays, as I am by festival programs, by reviews, by other forms of cultural programming. I sound off to my friends about programmers getting trapped by publicity machines, about independent curators doing the work of book publicists, about the booksellers going all out to promote the work of their former employees. 

Like Chayka, I’d prefer to consider the advice of a human tastemaker over an algorithmic recommendation – but I’ll reserve the right to bicker with them, to question their independence, and to assume that at the end of the day, a bookseller in whatever bookstore is still trying to sell me, the customer, a book. I have to say I found Chayka’s utopianism about human tastemaking a little strange in view of the furious public discourse of the past decade about the shortcomings of most systems of arbitration and distinction. If you call your curators gatekeepers, they’re a little less glamorous and the way they use their cultural power open to question. If you call them critics, they’re pariahs – but they might ultimately be more useful.  

Later in the book Chayka takes us to Loyalty Books in Washington, D.C., founded in 2018 by a woman named Hannah Oliver Depp. Loyalty Books he presents as an initiative that seeks to respond to algorithmic feeds and simultaneously to operate independently of them. Depp observed that the communities forming on Bookstagram and BookTok, especially queer readers and readers of colour, were not being welcomed by traditional bookshops. She sought to create a retail experience that was hospitable to those digital communities, firstly by providing playful spaces to shoot images and videos, and secondly by paying attention to the books and genres that are prominent on social media, especially romance and fantasy. The store, says Chayka, is ‘optimised for online content creation, because the store as well as the books must be distributed through algorithmic feeds just as much as function in physical space’. Depp tries to respond to whatever is hot on the feeds and to choose her own stock, to balance out what she sees as an often homogenous group of books that rise on social media; ‘the top-selling folks are straight white women writing somehow emotional books: self-help-oriented books, romance, or romance-adjacent’. Depp identifies supply as a challenge in keeping her shelves stocked with the books favoured by literary influencers. If a book blows up online, she can’t always get stock, and so loses business to Amazon, not because of its algorithm but because of its just-in-time distribution network. While Chayka is clearly setting up his reader to be appalled by the Amazon retail experience, his presentation of Loyalty Books is harder to parse. Depp is a business owner who is trying to have it both ways, to curate her own stock like a McNally Jackson, and to respond in an entrepreneurial manner to the trends that surface on social media. It seems like an impossible balance but at the time of writing, Loyalty Books is still in business and has opened a second store in Maryland. 

Loyalty Books is a tiny competitor to Amazon and is threatened less by the tech giant’s mastery of algorithmic recommendation than by its sheer scale, the ruthless efficiency of the Amazon supply chains. A friend recently posted some news about her forthcoming book, with a note to this effect: whatever you do, don’t pre-order my book on Amazon. She didn’t do this because she was angry about algorithmic recommendation. Her aversion is driven, as mine is, by Amazon’s predatory pricing, bad labour practices, and disgraceful climate footprint. Algorithmic recommendation might lie at the heart of Amazon’s success, but it is the policies and business decisions approved by Jeff Bezos, a human tastemaker in executive chairman form, that have really fucked up the book industry – that have made turning a profit challenging for the Loyalty Books and McNally Jacksons of the world, that have changed the stakes for independent publishers who want to take risks on new authors, that have trimmed sales margins in a manner directly affecting author incomes – all the while contributing measurably to global heating. 

Filterworld is Chayka’s umbrella term for the ‘vast, interlocking, and yet diffuse networks of algorithms that influence our lives today, which has had a dramatic impact on culture and the ways it is distributed and consumed’ and he gives voice to widely shared concerns about the pernicious effects of platform capitalism on the way audiences encounter culture and, increasingly, on the way artists make culture. It is a necessary topic and the reason the book has been so widely discussed is because it captures so well the uneasy ambience of contemporary culture. On Chayka’s account we are living in an epoch defined by a paradox: that ease of access to and distribution of culture has resulted in a narrowing of diversity and possibility in that culture. To make his case he proposes a peripatetic prehistory of the present impasse, one that includes case studies of cultural consumers, content producers and curators, all of whom are disillusioned with algorithmic recommendation.  

Woven into this narrative of the emergence of Filterworld is a personal history of social media use and cultural consumption. We sit with an adolescent Chayka on Messenger and in gaming forums, we keep him company when he stops downloading torrents and gets a Netflix subscription during the pandemic shutdowns. This centring of the author’s experience is a common device in contemporary narrative non-fiction and here it is unavoidable as Chayka is describing the soup we all swim in. Most of us – and more later on who Filterworld’s ‘us’ includes – consume at least some culture via platforms that deliver recommendations of books, music, TV, holiday accommodation, images, products that we might like based on our previous behaviour. Even consumers who don’t buy books via Amazon are familiar with the formulation: customers who enjoyed Boy Swallows Universe also purchased All Our Shimmering Skies. Chayka is persuasive and entertaining in demonstrating all the ways that algorithmic recommendation can get it wrong, and the way the experience of living in the Filterworld can be at once anxious and extremely boring. There’s some wonderfully revealing detail about the development of social information filtering and early sorting algorithms.  

Having traced a history of algorithmic recommendation, Chayka considers what remedies may be available to him and his readers. The project of the final chapters of the book is to ‘resolve the omniscient atmosphere of anxiety and ennui that algorithmic feeds have produced’. Chayka relates his own experience of an algorithm cleanse, also known as staying off the socials – particularly hard work for a journalist whose beat is digital culture. It’s a pretty light form of participatory reporting, so far as it goes, and hardly a radical cure. He also advocates for more intentional forms of digital media consumption, for audiences to access culture via platforms such as Patreon, Bandcamp, and the Criterion Channel that hearken back to the days of a more DIY internet and support creators directly. As Chayka is pessimistic about any prospect of structural change via regulation or collective action, it is the idealised figure of the curator who is a role model for audiences seeking to reclaim their agency. And so if there’s a future beyond Filterworld, it is fragmented, and individuals are left to their own fates: ‘we must also change our own habits, becoming more aware of how and where we consume culture rather than following the passive pathways of algorithmic feeds’.  

Algorithmic recommendation is a technology for delivering to users cultural content that their past behaviour suggests they will like. As users watch, scroll, listen, click, buy, and rate they provide behavioural data that allows for the refinement of recommendations, both in terms of establishing more detailed profiles of users and sorting content. Quantitative modes of cultural analysis displace the qualitative. Netflix may sort romcoms into a dizzying array of subgenres but it doesn’t make distinctions between or within categories. We spend time in Filterworld with Damon Krukowski, the drummer of an indie band called Galaxie 500, one of whose tracks, ‘Strange’ was disproportionately boosted by the Spotify algorithm relative to the rest of the band’s output. ‘Strange’ sounds nothing like anything else recorded by the band and was laid down, on Krukowski’s account, as ‘a loose parody of more popular music’. The working title for the track in production was ‘Heavy Metal Ballad’  and this is why, Chayka and Krukowski speculate, it hit the algorithmic jackpot and was recommended to listeners who were into heavy metal 80s ballads. The track delivered some modest income to Galaxie 500 but it was an empty, dislocated and unsatisfying experience for the artists, whose sense of their own creative project was distorted. None of this matters in the context of platform capitalism. A parody of a heavy metal ballad is a heavy metal ballad, as far as the algorithm is concerned. What matters is keeping users on the platform, tracking their behaviour, keeping them engaged. The currency of Filterworld is engagement; clicks, likes, follows, revenue. As Chayka puts it, ‘Social media has quantified culture into a banal set of metrics measuring views, click-throughs, and ultimately, purchase rates.’ 

This quantitative approach to culture was inaugurated well before the epoch of Filterworld and extends beyond the domain of arts and culture. Sales figures continue to lead assessments of cultural value and performance metrics prevail in every arena of public and private life. In his classic 1980 essay on the effect of television on American culture, ‘Within the Context of No Context’, George Trow describes and disdains a culture of counting: 

In the New History, nothing was judged—only counted. The power of judging was then subtracted from what it was necessary for a man to learn to do. In the New History, the preferences of a child carried as much weight as the preferences of an adult, so the refining of preferences was subtracted from what it was necessary for a man to learn to do. In the New History, the ideal became agreement rather than well-judged action, so men learned to be competent only in those modes which embraced the possibility of agreement. The world of power changed. What was powerful grew more powerful in ways that could be easily measured, grew less powerful in every way that could not be measured. 

Trow’s work is a lodestar for Chayka, and provides an epigram for Filterworld:  ‘The message of many things in America is “Like this or die”’. This sentence provided Christian Lorenzten with a title for his 2019 Harper’s essay on the fate of the book review in the age of the algorithm, in which he pre-empts several of Chayka’s principal concerns about the cultural impact of platform capitalism in a distilled form: ‘the basic imperatives of the review – analysis and evaluation – are being abandoned in favour of a nodding routine of recommendation.’ 

Chayka’s first book The Longing For Less (2020) explored the cultural pull of minimalism, and minimalist interiors are a particular preoccupation in Filterworld. He tracks a kind of degraded, modish minimalism through the cafes and Airbnbs he visits around the world, pointing to the trailing succulents, warm woods, just-so details that can make going to a cafe seeking to do business with millennial knowledge workers in Copenhagen pretty much the same as going to a cafe in Cinncinati. You know the place: aggressive claims about good coffee, white walls, plugs for laptops, a big awkward communal table. This aesthetic he calls Generic Coffee Shop and it distresses him greatly; ‘I felt bemused and dislocated by it, the way it feels slightly unreal to take an overnight flight and land in a different country.’ The cafes are asked to do a lot of work in Filterworld because, Chayka writes, they ‘turned out to be my canary in the coal mine for examining the impact of the Internet on cultural tastes and consumption patterns.’ The discussion of cafe aesthetics opens into a broader discussion of architectural flatness that careens from Thomas Friedman to Marc Augé via Gangnam Style, passing by the work of Rem Koolhaas. Chayka talks to cafe owners around the world about how algorithmic recommendation affects them and finds consensus on two points, firstly, that their business relies on being discoverable via platforms such as Yelp, Instagram and Facebook, and secondly, that their relationship to the platforms is unpredictable, unreliable and stressful. ‘I hate the algorithm. Everyone hates the algorithm’, one cafe owner tells him. I was left somewhat perplexed by Chayka’s attention to the cafes, unsure, in truth, of what the stakes were, especially in the context of Filterworld’s broader prescriptions about cultural curation. Does he want café owners to take aesthetic risks, diversify their interiors and possibly compromise their algorithmic standing? Or does he want coffee drinkers to stop relying on platforms to help them find somewhere to work at 11 o’clock in the morning in a new city? 

The emphasis on vibes is more egregious in Chayka’s discussion of Airbnb. Airbnb minimalism is the leading example of aesthetic normalisation driven by personalised algorithmic recommendation in Filterworld. Normal, writes Chayka, ‘is a word for the unobtrusive and average, whatever won’t provoke negative reactions. Whatever fits in that zone of averageness sees accelerated promotion and growth… while the rest falls by the wayside.’ For all Airbnb’s promise to help users enjoy a local experience, Chayka contends that in aesthetic terms, the short-term rental experience has been flattened, and it’s the same set of Ikea sofabeds and mugs wherever you go, an overnight version of Generic Coffee Shop. Instead of being an easy way for tourists to have an authentic experience of Barcelona, Airbnb stays yield a boring version of the same old aesthetic. Everything looks the same these days! One wonders whether Chayka has spent many nights in chain hotels. A reader might overlook this take as the kind of relatable chatter exchanged by knowing millennials, were it not such a recurrent theme of Filterworld. But to keep going back to the unpreposessing interiors of Airbnbs and to fail to look out the window and bear witness to the unbelievable impact of short-stay accommodation on housing in cities like Barcelona, Vancouver, Sydney, and small towns like Bellingen is to miss the forest for a single, mass-produced tree. In his consideration of popular tourist destinations, Chayka talks about the way ‘online ubiquity has a way of denaturing its subject’ but his focus is on the experience of the tourists themselves, with barely a gesture to the impact of platform-driven tourism on locals.  Aesthetic homogeneity is not inconsequential, but the distortions to housing markets all over the world wrought by Airbnb tourism are surely uglier. 

It’s hard to disagree with a lot of what Chayka has to say in Filterworld, but as I accompanied him on his melancholy whistlestop tour of contemporary culture, I found myself often questioning his emphasis on algorithmic recommendation at the expense of all else. For one, this approach risks letting technology companies off the hook, an effect redoubled by the mystification of algorithms. One of his interviewees tells him early on, ‘the Facebook algorithm doesn’t exist, Facebook exists. The algorithm is a way of talking about Facebook’s decisions.’ By the time I reached the end of Filterworld, I wished that Chayka had not taken this advice at face value and had spent more time talking about big tech. The trope of one all-powerful, all-purpose string of commands winking away on a server somewhere, tended by engineer-automatons, serves in Chayka’s work – and in so much cultural commentary on technology – as a metonym for an industry so fiendishly difficult to understand that it simply cannot be regulated. This is an effective and accessible way to dramatise the omnipresent and stupefying impact of technology companies such as Amazon, Alphabet (the parent company for Google, Youtube and much else) and Meta (which owns Facebook, Instagram and WhatsApp) on the production and distribution of culture. Consumers can seek to transform themselves and their habits – but they cannot hope to change the system. It’s not just a cause and effect problem – sometimes Chayka is so concerned with the aesthetic implications of algorithmic recommendation, such as the loss of the experience of bookstore browsing and all those boring interior spaces, that he glosses over the economic, social and environmental consequences of the decisions made by tech leaders. Chayka does provide his readers with a history. To advance a proper critique of algorithmic culture, though, requires more than a phenomenology. 

I was unconvinced by Chayka’s compartmentalisation of aesthetic experience – and I am also not persuaded by his arguments about aesthetic normalisation, which rely on unchecked assumptions about mass culture, about the phenomenology of reading, viewing, scrolling, about taste, and what it is that we seek in culture. His takes catch the reader’s attention – they sound like the beginnings of important, timely conversations – but often don’t reward close inspection. For example, he writes, ‘the culture propelled by Filterworld tends to be accessible, replicable, participatory, and ambient.’ This sameness, he argues, has a numbing effect and diminishes our capacity to be moved by culture. I can go along with this, even though I find myself asking, whose capacity to be moved? What is it to be moved by culture? People numbed their brain with bad TV in the days before Filterworld and critics, such as Trow or Adorno, sneered at them for being mesmerised into passivity. Since when did a recycled version of romantic transport serve as the benchmark for cultural impact? I want him to say more about this numbing as a form of disenfranchisement but he never does. It’s the next step that he takes in this passage that asks too much of me. In Filterworld, numbed by too much of the same thing, he writes, ‘our natural reaction is to seek out culture that embraces nothingness, that blankets and soothes rather than challenges or surprises, as powerful artwork is meant to do.’ Faced with assertions like this, I’m like a sessional tutor, forced to mark one too many essays on digital culture, possibly the worst kind of nitpicky reader and critic. Who says it’s a natural reaction to seek out culture that embraces nothingness? What is an artefect of a culture that embraces nothingness? Is it Beckett or is it Vanderpump Rules? Where does this hierarchy that values challenge over comfort come from? Who says that powerful artwork is meant to surprise? All these questions have answers and contexts – but you won’t find them in Filterworld. Insofar as Chayka conveys the bewildering contemporary experience of making and consuming culture, this is a valuable work of cultural journalism. What it lacks, however, is a critical approach to its subject. It is a work of synthesis rather than analysis.  

Filterworld addresses itself to people with some demographic contiguity to its author. Most books do. Chayka is a middleclass millennial culture writer with abundant cultural capital. He values books, music, film; he has time to spend on them, and some disposable income. He is educated, he travels, he greets the world with a combination of dread and entitlement. I’m a little older than he is, but on the whole, I recognise myself as part of the Filterworld demographic. We sometimes go to cafes to work and worry about spending too much time on our phones. The first person plural is, however, conspicuously unmarked throughout Chayka’s analyses and at no point does class really figure in either his diagnoses or prescriptions, or even the terminology of mass culture or middlebrow culture. Culture is ordinary, writes one of the great theorists of popular culture, Raymond Williams. Chayka wants it to be extraordinary, always. There’s a long tradition of impugning audiences for engaging with culture in the wrong way, for going to ghastly music hall performances, for corrupting their morals with rubbish novels, for playing videogames, for rotting their minds watching the tube. Chayka reproduces this anti-egalitarian cultural policing as if by reflex, creating exemptions for those readers knowing enough to seize agency, to become curators. Becoming a curator of your own cultural life sounds nice, it also sounds like something you need a lot of cultural capital to achieve. Chayka nods at Pierre Bourdieu in Filterworld but not when he’s asking his readers to transform themselves into better readers. It’s never examined why the movies screened on the Criterion Channel qualify as good culture and TikTok videos don’t. I want to sign on to Chayka’s appeal for a more qualitative approach to culture – at heart, I agree with him that metrics are killing culture – but I wish he’d been more careful in his thinking about popular culture, about the relationship between art and culture, that he had considered the vast body of work driven by, say, Stuart Hall and Williams, and been able to thus think about the relationships and ruptures between, say, Spotify lists and music hall, or Everybody Loves Raymond and Charlie D’Amelio. At the very least, watching the broadcast television of the late-90s and early-00s might have quelled Chayka’s declinist impulses. 

Another observer of cultural decline, Elizabeth Hardwick, made the following observation of popular reviewing in her landmark 1959 essay, ‘The Decline of Book Reviewing’: 

Everything is somehow alike, whether it be a routine work of history by a respectable academic, a group of platitudes from the Pentagon, a volume of verse, a work of radical ideas, a work of conservative ideas. Simple ‘coverage’ seems to have won out over the drama of opinion; ‘readability’, a cozy little word, has taken the place of the old-fashioned requirement of a good, clear prose style, which is something else. All differences of excellence, of position, of form, are blurred by the slumberous acceptance. 

This sounds a lot like the flattening that Chayka attributes to algorithmic culture and chimes with Carl Wilson’s recent pithy response to Filterworld in Bookforum

I wonder if Chayka is experiencing the worst of the algorithms partly because he’s just now moving into middle age. That’s the point when most people all of a sudden ‘discover’ that something has gone wrong with pop culture.  

Filterworld closes with calls to action for cultural consumers who wish to evade the flattening effects of algorithmic recommendation and thereby re-enchant their encounters with culture.  

To resist Filterworld, we must become our own curators once more and take responsibility for what we’re consuming. Regaining that control isn’t hard. You make a personal choice and begin to intentionally seek out your own cultural rabbit hole, which leads you in new directions, to yet more independent decisions. 

Take responsibility for your choices? This sounds like a neoliberal solution to the cultural crisis of neoliberalism. This is probably too swift a dismissal, and yet it’s chilling to see the language of personal responsibility deployed as a response to what is clearly a structural, epochal issue. Curation, as Chayka observes, has become a ubiquitous synonym for a kind of knowledgeable cultural sorting. Editors curate homepages, influencers curate Christmas gift selections; sometimes it can be difficult, in a digital media environment, to distinguish between influencers, curators, editors and publicists, all of whom appear to be sorting goods for sale. Chayka seeks to restore some integrity to the term curation which is, etymologically, a form of care-taking; he presents it as a way for audiences to take more care with culture. ‘The slow process of curation,’ he writes, ‘works against the contextlessness, speed and ephemerality that characterises the Internet.’ I appreciate Chayka’s efforts to activate a kind of artisanal imaginary, his advocacy for a thoughtful and autonomous approach to culture. And yet, even in Chayka’s enlivened usage, curation seems to involve more often than not a form of intentional picking and choosing rather than evaluation and analysis.  

A curator might be an active consumer of culture, but they’re still, following Chayka’s thinking, a consumer. Perhaps we need to relinquish consumption as the guiding metaphor for encounters with art and culture in favour of – what? Participation? Rapture? Distraction? Invigoration? Transformation? Perversion? Enfranchisement? Perhaps we need to let art and culture revivify our engagement with the social and economic realities of this cruel century, rather than annexing aesthetic experience as Chayka so often does. Every now and again Filterworld tiptoes near promising terrain: that in restricting the autonomy of artists and their audiences, algorithmic recommendation inhibits the liberatory, radical and flagrantly democratic possibilities of culture, that artists themselves might show us how to tear down the structures of platform capitalism. And then it retreats, to an altogether milder model of emancipation via the algorithm cleanse, and the rather diminutive heroism of the curator. I wonder whether a more robust role model for audiences – readers – seeking to escape algorithmic ennui, to resist ‘slumberous acceptance’ might not ultimately be the critic.