“Humans are distinguished from other species”, says Peter Thiel, one of Silicon Valley’s high priests, “by our ability to work miracles. We call these miracles technology.” Thiel inadvertently touches on a pervasive paradox: we see ourselves as both the miracle-makers of technology and the earthly audience, looking on in wonder. But if the miracle was once the automobile, the modern equivalent of the “great gothic cathedrals”, in Roland Barthes’s famous formulation, now it is surely the internet: conceived by unknown forces, built on the toil of a hidden workforce, and consumed more or less unthinkingly by whole populations. The internet’s supposed immateriality masks not only the huge infrastructure that sustains it, including vast, heavily polluting data centres, but also the increasingly narrow corporate interests that shape it and, in turn, us – the way we think, work and live. Algorithms are at the heart of this creative process, guiding us through internet searches and our city’s streets with a logic steeped in secrecy, filtered down from above – namely, the boardrooms of the Big Five: Amazon, Apple, Alphabet Inc. (the parent company of Google), Facebook and Microsoft, those companies that have come to dominate the digital realm.
Subscribe to the weekly TLS newsletter
Ed Finn’s What Algorithms Want: Imagination in the age of computing is an attempt to demystify these “quasi-magical” algorithms. The title could suggest a dating manual from the not too distant future – Spike Jonze’s film Her comes to mind – and to an extent the book is not far off. Finn writes to ground these elusive entities in their material existence, in the hope of making our coexistence happier and more creative. His ambitious attempt to draw together all the different algorithmic platforms into one cogent tale might occasionally falter – they are simply too varied for a book this slim – but the frenzied focus, from Google and Facebook to Airbnb, Uber and gaming, at least gives a fitting sense that, as Finn puts it, “algorithms are everywhere”. The word itself, he explains, derives from the ninth-century mathematician Muhammad ibn Mùsà al-Khwàrizmì (latinized as Algoritmi), a Persian who laid the foundations – and coined the term – for algebra. “Algorithm” gradually came to signify any mathematical process that navigated its way through data sets in a given way: “a sequence of tasks to achieve a particular calculation or result”, whether a square root or, today, a Google search. Indeed, computer engineering’s adoption of the term artfully placed programming in the cold, scientific current of mathematics, where complex problems have singular solutions, free from subjective bias. This imbued the algorithm, at least initially, with an aura of impartiality. “Algorithms are the computer processes and formulas that take your questions and turn them into answers”, Google explains. Based on our past online behaviour, they are also enlisted to suggest us potential partners, summer reads and news stories. In some cases, algorithms are entrusted with determining a criminal’s likelihood to reoffend, selecting employees to sack, navigating the stock market and even writing the news stories themselves. Add to this growing trust the secrecy that often surrounds their method of calculation (the preservation of “trade secrets” ostensibly trumping that of transparency, even in America’s criminal justice system), and algorithms become, as Finn argues, “mathematical oracles”.
Yet these supernatural capabilities – sifting through enormous data sets within a fraction of a second – distract from all too human limitations. Their goals and guiding principles are set by a coterie of engineers and executives united, for the most part, by shared backgrounds and beliefs: either a geeky fondness for computer code or an infinite ambition for what it can achieve, or both. Among the top ten major tech companies, women make up less than 30 per cent of the workforce – in technical roles, this number drops to 18.3 per cent. While black people and Hispanics make up, respectively, about 13 and 16 per cent of the total American workforce, according to a recent report in Wired, about 2 per cent of the tech workforce is black and 3 per cent is Hispanic. What is more, with data inevitably drawn from the past, the algorithm’s logic is imbued with the biases, flaws and injustices that come with it. In this context, it is no surprise that their impartiality often descends into its opposite. Automated decision-making becomes automated discrimination: prison mugshots are more likely to show up when someone searches for images of black teenagers; prominent job openings are more likely to be shown to men; “the doctor” is translated into Spanish as “el doctor” not “la doctora”; a convicted black man is declared more likely to reoffend.
As the designers typically work for private companies, these problems only become more pronounced. Earlier this year, after an eight-year investigation, the European Union concluded that Google was abusing its market dominance by manipulating search results to suit itself, warranting a record-breaking €2.42 billion fine. Algorithms were not simply taking your questions and turning them into answers, as the company claimed; they were taking your questions and, wherever possible, prioritizing Google’s own products and services, thereby converting questions into maximum cash. Google’s paradoxical response was both to appeal against the fine and address its underlying complaint: in Europe, Google now allows other firms to bid for the top-of-the-page slots in their “Google shopping” box – a profitable solution to the problem. Jonathan Taplin’s Move Fast and Break Things: How Facebook, Google, and Amazon have cornered culture and what it means for all of us, which was published before the EU’s decision, is in many ways the opposite of Finn’s more curious inquiry. Taplin’s message is urgent and unremitting: Silicon Valley’s dominance is endangering democracy – socially, culturally, economically and politically – and at the very least, these companies should be subject to greater regulation.
The scale of their supremacy is remarkable. Google controls more than 90 per cent of search-engine traffic in Europe, and 88 per cent in America; YouTube, also owned by Alphabet, is the largest music-streaming site in the world; Amazon takes more than half of every new US dollar spent online; Facebook, having swallowed competition like Instagram and WhatsApp, holds 77 per cent of all mobile social media traffic, and recently reached 2 billion members. With Apple and Microsoft, these are now the five largest companies in the world – hence the group moniker – with a collective net worth of $3 trillion. Together they spend more on government lobbying than the five biggest banks or the five largest oil companies. “Not since the turn of the twentieth century,” Taplin writes, “when Roosevelt took on the monopolies of Rockefeller and Morgan, has the country faced such a concentration of wealth and power.” All this, he adds wryly, on the back of an invention that was supposed to “eliminate the gatekeepers”. Not that Silicon Valley sees anything sinister in all this. As Taplin puts it, they have “a very high opinion of their place in history”. According to Sergey Brin, Google’s co-founder, the company aims to be “the perfect search engine . . . like the mind of God”. Earlier this year, Mark Zuckerberg even seemed to suggest that Facebook should emulate, if not replace, the Church as the bedrock on which communities are built. “Mark Zuckerberg will never be my pastor”, one reporter wrote in Christian Today.
For all their missionary zeal, Google and Facebook remain predominantly advertising companies – and their algorithms are ordered accordingly. They give their services away free, and make billions collecting data and selling it on. And like the advertising industry itself, their aims are riven with contradictions. Claims to cultivate the unique character of the individual, with increasingly personalized products, conflict with the greater goals of consumerist conformity and capturing user attention: self-expression can run free, so long as it plays out along predictable, profitable lines. Hence, for example, Facebook’s increasing move towards auto-play video on its newsfeed, and the auto-queue on YouTube that sends you to a song with more views. In terms of “innovation” these companies present themselves as radical pioneers – they move fast and break things, as the mantra goes – but in practice, this only veils the homogenizing effects of their services. The music industry is an emblematic case: when the algorithms of streaming services push us towards what is already popular, a trade where 80 per cent of the revenue derives from just 1 per cent of the product is the result.
As Taplin stresses, however, perhaps the most worrying side of Silicon Valley’s self-aggrandizing aspirations is neither the money the companies make nor even the consequences so far. It is their contempt for whatever stands in their way. Democratic processes, the state, the public, tax, a basic sense of social responsibility, all are seen as so much baggage on a holy quest to better the lot of humankind. Ayn Rand’s refrain in The Fountainhead – “Who will stop me?” – has become a guiding principle. The Big Five were members of the American Legislative Exchange Council, a shadowy organization made up of the biggest corporations dedicated to advancing the free market, until a public outcry forced them to quit. While their founders and executives deliver sermons on, say, the need for a (tax-funded) universal basic income, the companies themselves excel in avoiding tax. In 2014, Facebook’s UK offices managed to pay £4,327 in corporation tax, less than the average British worker pays; in the same year, bonuses to its UK employees reached a total of £35 million. In 2013, Google paid £20.4 million in UK tax on a revenue of £3.8 billion. Taplin also points to a tellingly popular “sea-steading” movement, which sees Silicon Valley dream of “building an artificial island in the middle of the sea, not hosted by any nation-state” for the purpose of innovation. Thiel, one of its main funders, pictures it as an “escape from politics in all its forms” (and, one imagines, tax). They recently reached an agreement with the French Polynesian government to take over a portion of its territory instead. Meanwhile Larry Page, Google’s co-founder and Alphabet’s CEO, echoes the idea that governments should “set aside a small part of the world” for them alone – the company is now looking for land to build its own city; the venture capitalist Tim Draper has suggested that Silicon Valley become its own state. Add to this these companies’ vast, self-contained campuses, and one is left with the clear sense of something like a modern monastery, cut off from society but with the weight of humanity’s hopes on its shoulders: Silicon Vatican.
With all the data they collect, they are certainly gaining a God’s-eye view. “We don’t need you to type at all”, said Eric Schmidt, former Google CEO and Alphabet’s current executive chairman. “We know where you are. We know where you’ve been. We can more or less know what you’re thinking about.” The etymology of the word Vatican is, incidentally, vatis, meaning prophecy or divination, and for Silicon Valley, prophecy (in the name of profit) is the name of the game. The more they can predict our actions, the more they can be monetized, and so every click, message, scroll and GPS co-ordinate is recorded and added to personal profiles. Of course, prediction works best when actions are pre-determined, so the internet is shaped – and our behaviour reshaped – with these interests in mind. Several studies have shown that we reveal far more details about ourselves online than we do in conversations offline. This isn’t an inevitable side effect of the internet: it is a reflection of corporate aspirations. “The days of you having a different image for your work friends or co-workers and for the other people you know are probably coming to an end pretty quickly”, Zuckerberg says. Apparently, “having two identities for yourself is an example of a lack of integrity”. Social media sites thus define sharing more and acquiring more “friends”, “followers” and “likes” as signs of success and authenticity. So when you change your profile picture on Facebook, for example, an algorithm will make this appear higher on the newsfeed, aiming to generate the greatest number of likes because it sees this as a moment when you will be seeking validation, and Facebook wants you to do it again. Refrain from sharing and Facebook might send you a disparaging notification: “You last updated your profile two weeks ago”. Share, and your progress will be praised: “You’ve shared 2 days in a row and your friends are responding”. Framed like this, our online activity clearly is a form of labour that Facebook seeks to discipline into our daily habits. And it works: in 2014, Taplin writes, “Facebook’s 1.23 billion regular users logged into the site for seventeen minutes each day . . . that’s more than 39,757 years of time collectively spent on Facebook in a single day. That’s almost fifteen million years of free labor per year”. Today, Facebook has over 2 billion users spending an average of fifty minutes on the site, leaving other forms of capitalist production to look oddly quaint. “Karl Marx would have been totally mystified”, Taplin reflects.
Zuckerberg, for his part, has claimed that any notion of a conflict between the interests of Facebook’s users (“our community”) and the company’s need to interest advertisers is “the most ridiculous concept”. Tracking every digital activity, including a user’s browsing behaviour beyond Facebook, supposedly serves a higher end: “I’m also curious about whether there is a fundamental mathematical law underlying human social relationships that governs the balance of who and what we all care about”, Zuckerberg mused in 2015. “I bet there is.” Earlier this year, a leaked Facebook document revealed a much baser project: the company was pitching to advertisers its ability to identify when young users were feeling “worthless”, “insecure”, “stressed”, or “defeated”, based on their online activity. Not a mathematical law, but those “moments when young people need a confidence boost” that the advertising industry relishes. A more recent investigation by ProPublica found that Facebook was letting advertisers target individuals who had liked groups defining themselves as “jew-haters”, the “Ku-Klux-Klan” and “kill Muslimic groups”. (Facebook has since removed this function.) One woman also recounted going to Google with a genuine concern: “How do you know if you have a drinking problem?” A few hours later, she was greeted on Facebook by an advert for her local liquor store.
Such stories are only shocking because of these companies’ lofty claims to be doing good. In the broader context of advertising, they make perfect sense. “It has always been in sellers’ interests to know their customers better”, Mark Bartholomew explains in his new book Adcreep: The case against modern marketing. All that has changed is how well they can do it. With a wide range of case studies at hand, Bartholomew places these increasingly invasive tactics in their historical context. Like Taplin, he suggests that we take inspiration from the past. Previous legal successes, he argues, such as the victory against subliminal advertising in the 1950s, “reveal the possibility of resistance to marketing encroachments”. But he is dismayed by today’s “legal lethargy”. “At a time when a panoply of new marketing techniques is changing human behaviour and eroding consumer agency,” he writes, “the legal system has stood still.” More than Taplin, Bartholomew recognizes that a change in the law will require a change in our mindsets. We need to show “a willingness to protest”, he says, by “using anonymous search engines, or taking sabbaticals from social media”. Yet leaving aside the effectiveness of such strategies, resistance seems unlikely, partly because of the intentionally addictive appeal of the technologies – something Mark Bartholomew fails to mention, contra Jonathan Taplin – and partly because, as with any addiction, we isolate our enjoyment – or need – from the darker consequences. When Facebook crashed in some US cities during the summer of 2014, many people called 911.
Something has to give. As these companies take increasingly detailed data to deliver increasingly predictable and profitable versions of ourselves, the more time we spend in the presence of their projections, the more these projections risk becoming the real thing. Our character and culture are being remade to the rhythms and requirements of Silicon Valley’s financial interests. Many of the infamous problems – fake news, filter-bubbles, invasive advertising – are not so much technological flaws as extremely lucrative features. And while calls for greater regulation are rising on both sides of the political divide, it’s unclear what shape this regulation would or could take to be effective. Our enchantment with digital devices has now elevated innovation to an end in itself, shrinking our horizons to the size of the consumer self. Improving the world and updating our operating system are reduced to equivalents, often leaving society’s hierarchies of power, race, wealth and gender for the most part intact. It is as if, under the guise of chasing the future, Silicon Valley sets the present on an endless loop, a hypnotic cycle that will require more than new regulations to break.
Subscribe now to the TLS and get the best writing on big books and big ideas from only £1.50 or $2.40 per weekSubscribe