The Long Review: Two new books attempt to examine how platforms and their algorithms vine into our desires, preferences and self-conception – and the answer is always messier than it first seems
Kyle Chayka’s Filterworld: How Algorithms Flattened Culture begins with the story of the Mechanical Turk, an eighteenth-century device that could play and win games of chess to the awe and delight of the Habsburgian court. It claimed to rely on gears and belts to react to the moves and plays off its opponents – a technological feat akin to magic. However, the Mechanical Turk had a secret: its insides hid a human chess master who controlled the whole apparatus to give the illusion of an intelligent machine.
Chayka compares the Mechanical Turk to the algorithms that influence our lives through social media and consumer platform feeds. Wherever there is an email, search result, dating prospect, song, video, TV show or book in need of sorting, algorithms step in to decide what rises to the top. These algorithmic interventions are often imbued with magical powers, as if they are able to divine a user’s needs and summon a solution when, in fact, they metabolize the data we input through our behaviour to best guess what it is we’d like to see (or buy).
Chayka chose to tell the story of the Mechanical Turk because he finds that like our ‘diffuse network of algorithms’, the Turk represents ‘the human lurking behind the facade of seemingly advanced technology’. He goes on to talk to Nick Seaver, a professor at the department of anthropology at Tufts University, who tells him that ‘the algorithm is metonymic for companies as a whole’. In other words, the infamous ‘algorithm’ named in headlines and viral tweets about targeted advertising or magical ‘For You’ pages is actually a series of algorithms, plus the many interface design choices and business practices that social media platforms and content streamers build and maintain as part of their product. Seaver continues: ‘The Facebook algorithm doesn’t exist; Facebook exists. The algorithm is a way of talking about Facebook’s decisions.’
By beginning the book with the story of the Mechanical Turk, Chayka aligns himself with Seaver and other critics who reject the mystification – glorification or vilification – of the algorithm as a technology. He recognises how popular talk of algorithms falls prey to the obscuring powers of technology, a skepticism of ‘the ability of such devices to deceive us about the way they work’, and how that trap is itself a product of the lack of transparency on behalf of the companies that use them.
However, just one sentence after Seaver’s quote, Chayka loses focus. He adds that the technology itself ‘is not at issue’, and that the cultural flattening he bemoans in the book’s title is because we’ve outgrown these algorithmic recommendations and are now ‘alienated by them’. Throughout the book, Chayka struggles to pair his Mechanical Turk metaphor with a bone-deep understanding of what it means for people – from engineers to executives – to be the designers of the very system of recommendation he accuses of flattening culture.
In Filterworld, Chayka relies on the metonymic algorithm to step in on behalf of deeper explorations into the myriad actors, motivations and incentives that contribute to the oh-so-offensive Generic Coffee Shop and Airbnb Tourism – products whose offerings and aesthetic are streamlined and shaped by algorithms. Everyone buys, does, looks at and identifies with the same handful of consumer goods and experiences because of such algorithms, Chayka says. An algorithmic ‘cleanse’, Chayka then decides, might be the first step in wrestling back some control and taking ownership of his personal taste. The cleanse is like a stress-test for his agency as an internet user, but his use of an almost insidious passive voice belies the result. During his cleanse, Chayka learned how his brain ‘had been completely trained on the dopamine hits of social media attention’. Elsewhere, ‘consumers are herded into a handful of massive platforms’ and the desires and preferences of the generic global consumer ‘are molded by the platforms they use’ (italics mine). In Taylor Lorenz’s Extremely Online: The Untold Story of Fame, Influence, and Power on the Internet, the word ‘force’ similarly steps in to gloss over ambiguity. In her introduction, Lorenz declares that her ‘social history of social media’ is ‘about a force and an industry that is upending legacy power’. Twitter’s arrival was ‘a cultural force’. Just as in Filterworld ‘algorithmic recommendations have become a force’; from Twitter to Instagram to TikTok, algorithms are an ‘invisible force’. How exactly it is that these platforms and their algorithms vine into our cultures, desires, preferences and self-conception and tangle us into knots of ‘algorithmic anxiety’ remains an unexplored mystery.
This kind of overreliance on metonyms and obfuscating language is an unfortunate trend of popular internet culture writing. To the credit of many writers and critics, the information needed to appraise things like recommendation algorithms, feed changes or community guideline implementations are too often concealed as trade secrets. This absence of crucial information can have a stifling effect, but internet culture writing nonetheless bloomed as a journalistic beat dedicated to covering the cultural effects of the business and technology of the internet.
Internet culture writing lives primarily in the culture, tech, youth culture or opinion sections of mainstream American and British digital media – Chayka and Lorenz are columnists at the New Yorker and The Washington Post, respectively. The ‘internet’ part generally engages subjects related to social media, creator economy, digital culture and life online. The ‘culture’ pertains to pop culture, consumer trends and moments of overlap with fashion or entertainment. At its best, internet culture writing has the potential to offer readers an interdisciplinary meditation on the ways life on- and offline influence each other. Internet culture writing is as good a space as any to consider how writing as a technology, genre as ideology and culture as a rhizome, complement the way life online electrifies shifting values and evolving categories. This kind of writing can not only narrate, but echo the unidirectional nature of the interactive internet. It can be writing that, even on the most granular level of language, bears the consequences of the changes it describes. It’s not just social commentary, it’s cultural criticism, entertainment and technology reporting, and media theory in-the-making.
In Filterworld, these obfuscations seem to be mostly an accident of the writing itself. The use (or abuse) or the passive voice produces a sense of doomerism, a neoliberal helplessness where companies can’t change, capitalism rages on, nobody is unique and the user (and writer) is helpless in it all. ‘We were stuck in an algorithmic flow,’ Chayka writes, ‘driven by whichever variables it was programmed to seek.’
In the book New Dark Age: Technology and the End of the Future (2018), artist and technologist James Bridle cuts through the design and forfeits the helplessness to pursue a closer look at the software and how it infects our lives. They look at trade algorithms with names like Ninja, Sniper or The Knife, ‘designed by former physics PhDs to take advantage of millisecond advantages’ and capable of ‘eking out fractions of a cent on every trade, and they could do it millions of times a day’. Bridle’s take is historically and technically informed, clarifying how and why algorithms work the way they do: ‘Smart or dumb, emergent or intentional, such programs and their usefulness as attack vectors are escaping the black boxes of stock exchange and online marketplaces and entering everyday life.’
While Chayka’s Filterworld stops at the acknowledgement of how humans make algorithms, in her 2018 book Algorithms of Oppression, Safiya Umoja Noble looks at how corporate and engineering cultures shape software and its implementations. She writes that ‘algorithms are a fundamental invention of computer scientists who are human beings – and code is a language full of meaning and applied in varying ways to different types of information’. It’s not just the fact that people employ the algorithm, it’s also the ideology baked into its code. Noble gets as close to the metal as possible to pinpoint the problem of algorithmic recommendation and map out better solutions.
For Chayka, the algorithms that cause the cultural flattening are simply flawed recommendation algorithms, the answer to which is fewer algorithms via regulation and better kinds of recommendation via museum curators and radio DJs, whom he idealises for their abilities to cultivate ‘an atmosphere of trust in which you can take in new cultural items’. But Noble’s understanding of who these algorithms actually serve – and how – explains the root of the issue differently: ‘Google creates advertising algorithms, not information algorithms’. The same scrutiny should be applied to radio DJs and museum curators – or, for that matter, any entity with the power to platform and filter what we see on- and offline.
It’s true that the public is limited in what we can know about the technologies and business decisions that shape our lives online, and it’s true that this limits the way we can write about them. But in 2018, Noble and Bridle found ways to write about that obscurity, its design, its effects, its knowable causes and many impacts. They refused to be passive in both their voice and their research and as a result, their work remains relevant.
On the internet we’re users, subscribers, creators, commenters, customers and commodities. Algorithmic feeds collect and metabolise hoards of our personal data, so no matter what we’re doing – posting or scrolling – we are always doing two things: using and being used. As such, it’s never as simple as traditional media and entertainment writing or cultural criticism, where the creating side is clearly demarcated and the audience can be placed at a distance. The work of internet culture writing is to outline how we all vacillate in our roles as subjects and objects.
Lorenz’s book, which she calls a ‘social history of social media’, casts creators – power users who make careers out of contributing to platforms – as the heroes in her retelling of how social media became what we know it to be today. ‘Over the past two decades,’ she writes, ‘an unprecedentedly innovative community has managed to refresh itself with every iteration of the social media landscape.’ In Lorenz’s retelling, tech CEOs and brand representatives can’t recognise the contributions and value of the creator economy fast enough. ‘From the first amateur blog to the newest TikTok sensation, it has been users and those in their periphery who’ve brought the creative energy, the tech companies rising around them, fueled by the rich content and collective attention.’ To sustain her thesis, Lorenz treats consumers and producers as two separate categories that only just started to intermix in the last 20 years.
Lorenz argues that platforms like Vine, especially, would’ve found more enduring success had they worked with rather than against their creators. In the case of Vine, ‘creators weren’t just angry about monetization’, she explains. ‘They were frustrated that the app seemed to have no interest in building features that catered to power users like them.’ Chapter after chapter, creators are disillusioned by their success, while companies see them as little more than very dedicated users. On their side, the creators seem stuck in their self-perception: they don’t exactly work for the platforms, but they bring a lot of value to them.
When she writes about the mommy bloggers of the 2000s, Lorenz credits WordPress, the blogging and content management service, for turning ‘media consumers into media producers’. However, a more interdisciplinary and historically attuned approach can be found in Lisa Nakamura’s 2007 book, Digitizing Race: Visual Cultures of the Internet. Instead of baptising the mommy bloggers as users-turned-creators, Nakamura finds their spaces to be a prime example of how the social internet collapses these categories. She describes how pregnancy puts women in positions where they interact with sonograms and other medical technologies that create images of them as background for the newly prioritised foetus. ‘The internet,’ however, ‘provides a space in which women use pregnancy Web sites’ modes of visuality and digital graphic production to become subjects, rather than objects, of interactivity.’
We are at the point where describing the move between user and producer as a unidirectional transition from one category to another will no longer suffice. The general public is, by now, mostly aware of how any time spent online is time spent generating value for the companies that own the platforms – from the influencer to the hate-follower clicking through the affiliate link. In 2018, Noble quotes researchers from the University of Maryland to characterise this collapse as a ‘trend toward unpaid rather than paid labor and towards offering products at no cost, and the system is marked by a new abundance where scarcity once predominated.’
‘We all have the memory of a goldfish now’, Lorenz said in an interview. ‘I think it’s so important to remember history.’ That history, however, is about more than individual creators and the companies that refused to reward them. After all, there is more to history than main characters and chronology of events: it is when and how the very subjects and ideas at the core of narratives came to be. The future of internet culture writing should, at the very least, harness the abundance of the web to build on and iterate its own history. A history of the internet – or any history for that matter – is deep and messy, a cacophony of actors, institutions, patterns that stretch through time and actions that move across contexts.
The internet is a lot of things. In fact, it might be too many things. From rare-earth minerals in processor chips and fibre-optic cables under our streets to the sleek Silicon Valley workers, defence ministries and executive decision-makers, our understanding of the internet is a baroque thicket of information on the move. Internet culture writing can surely fight the opacity that conceals abuses of power by connecting moments in culture to strains of ideology and changes in technology. But more interestingly, I look forward to seeing how internet culture writing can help us metabolize changes in hierarchies and cultural values – and transform how we write in general.
Michelle Santiago Cortés is a writer and critic living and working in Brooklyn, New York. She writes essays about digital culture and life online, as well as works of criticism about art and technology.
Filterworld: How Algorithms Flattened Culture by Kyle Chayka. Doubleday (hardcover), $28
Extremely Online: The Untold Story of Fame, Influence, and Power on the Internet by Taylor Lorenz. WH Allen (softcover), £16.99