The “dust of exploded beliefs,” the English aphorist Geoffrey Madan once wrote, “may make a fine sunset.” We’re beginning to see that glow over the internet which, if you count back to the design phase in the autumn of 1973, is now over four decades old.
From the moment the internet first opened for semi-public use in January 1983, it evoked utopian dreams. It was easy to see why. Cyberspace—the term coined by the novelist William Gibson for the virtual space behind the screen—really did seem to be a parallel universe to “meatspace,” the term invented by Grateful-Dead-lyricist-turned-essayist John Perry Barlow for the messy physical world that we all inhabit. Cyberspace in the 1980s was a glorious sandpit for geeks: a world with no corporations, no crime, no spam, no hate speech, relatively civil discourse, no editorial gatekeepers, no regulation and no role for those meatspace masters whom Barlow called the “weary giants of flesh and steel.”
But then, gradually, the internet was commercialised and those two parallel spaces merged to create our networked world, in which the affordances of cyberspace combine with surveillance and corporate control. Of course, the internet has brought huge benefits in terms of access to information and efficiency of communication: try imagining our home or work lives without it. But there are serious worries. The online world is populated by several billion mostly passive addicts of devices, apps and services created by a handful of corporate giants. Prying governments and giant companies have acquired the capacity to surveil our every move, both on the internet and, now that so many devices have built-in GPS, in the real world too. Through their ability to monitor our searches these companies—as well as the governments they co-operate with—are able to see our innermost thoughts and desires. (Yes, even our desires: what people search for on Google is incredibly revealing.)
It all creates the potential for unprecedented manipulation, and—rather suddenly—worries are piling up about how that network technology is disrupting our society, warping our children’s development, our politics and our lives. Even the digital evangelists are having second thoughts. In November 2017, Sean Parker, the first President of Facebook, accused the social network of exploiting “vulnerability,” with “God only knows” what effects on “our children’s brains.” In the same month, Chamath Palihapitiya, a former Facebook Vice-President, said social media firms had created “tools that are ripping apart the social fabric of how society works.”
The answer, as Hemingway said of going bankrupt, is first slowly and then very quickly. The tipping point came in 2007 with the launch of the iPhone, a product whose brilliant marketing and slick design would make the smartphone mainstream. Most people now go online using such a device. This is significant because, unlike PCs and laptops, most smartphones are closed devices, tightly controlled by manufacturers and network operators. The switch to mobile brought a sudden increase in corporate power.
The potential for corporate capture of the internet, however, was always there, thanks to a wider set of characteristics of digital technology: zero marginal costs (it costs Google next to nothing to register one more Gmail account); the power of network effects in cyberspace; the dominance of “power law” distributions, which lead to a small number of firms or actors dominating while everyone else languishes in the so-called “long tail”; technological lock-in, where a proprietary technical standard becomes the industry standard; and, most novel and most disturbing, the capacity for comprehensive surveillance.
Listen: John Naughton discusses the web of control with Prospect editor Tom Clark
The first four of these properties tend to enable winner-takes-all outcomes. And the last has enabled not only the ubiquitous state surveillance which whistleblower Edward Snowden exposed to an alarmed world through his revelations about the National Security Agency in the United States, but also a new business model for internet companies:
“surveillance capitalism.” In this arrangement, the company provides free services in return for unlimited access to users’ personal information and data trails, which are then refined, packaged and sold to advertisers.
Perhaps we should not be so surprised that the techno-visionaries’ utopian dreams have given way to a profiteers’ pleasure garden. This has been the pattern with every breakthrough in communications technology. The legal scholar Tim Wu showed in his book The Master Switch (2011) a magisterial history of communications media in the 20th century, that each technology—the movies, the telephone, broadcast radio and television—went through a similar cycle. It would start out gloriously chaotic, creative, open and free, but was eventually captured by corporate interests, sometimes abetted by the state. Wu’s big question was would the same happen to the internet?
We now know the answer. The network—and the tech industry—is dominated by five huge corporations: Apple, Alphabet (the owner of Google), Amazon, Microsoft and Facebook. Four of them—Apple, Alphabet, Amazon and Microsoft—are the most valuable companies in the world. Valuations bob about, but at some recent moments Facebook has been in fifth place. All five, then, are giant companies, but there are crucial differences between them. Whereas Apple, Amazon and Microsoft are extraordinarily disruptive in all sorts of ways (for just one example, see Houman Barekat on publishing) they are also recognisably conventional businesses which provide goods and services to paying customers. The other two, by contrast, are masters of a new surveillance capitalism. The billions of people who use Google and Facebook are not their customers. Instead, advertisers are—which has given rise to the mantra “if the service is free, then you are the product.”
Many of the grumbles about Apple, Amazon and Microsoft are much the same as those that were heard about monopolists of the past. But the same charges are not, perhaps, heard quite so often, or pressed quite so hard, against Google and Facebook. This is not because they are any less dominant in the sectors they operate in than the other three. Between them, the pair account for over 70 per cent of data traffic on the internet; they also take two-thirds of US digital advertising revenues. Google’s share of the search market ranges from 80-95 per cent, depending on the territory. Over 80 per cent of smartphones use Android, Google’s mobile operating system. That sounds pretty close to a monopoly.
Having users who are not customers is not the only thing these two companies have in common. Both also have two-tier shareholding structures, which means that a small number of founders and senior executives control them. The two companies are also alike in the dizzying scale of the data they process, and the sheer number of data traces that their users leave behind. Google, on average, handles 40,000 search queries every second (that’s 3.5bn searches a day or 1.2 trillion searches annually). In any given minute on Facebook, 136,000 photos are uploaded, 510,000 comments are posted and there are 293,000 status updates. Gmail has a billion users worldwide. Facebook has over 2.2 billion users—that’s 56 per cent of all internet users and 30 per cent of humanity. Half of these people use its various services every day, spending an average of 50 minutes in its clutches. Rarely in the field of commercial endeavour can one company have taken command of such a large chunk of so many lives.
It is the extraordinary levels of user engagement that allows Google and Facebook to operate as an unprecedented kind of commercial entity. A new beast has arrived in the capitalist jungle and we have no idea how to handle it. There are aspects of the companies’ influence that were apparently not foreseen by the firms themselves either, including the way that their advertising systems would allow pressure groups, parties and governments to covertly target voters with precisely calibrated political messages. It hardly matters that this influence wasn’t sought by either firm, when liberal democracy is nonetheless being shaken by the consequences.
In both the US presidential election and the Brexit referendum, tech-savvy political actors, potentially including hostile foreign powers, were able to use online tools to manipulate public opinion. The exact practical consequences in 2016 can be debated, but the potential for manipulation is plainly there—and the effect on the political culture is already evident. Campaign messages can be directed towards receptive ears, and directed away from those that may have been alarmed by them. Thus it is that the forward march of technology leaves politics stuck behind, in an echo chamber.
The heart of darkness lies in the vast amount of information they have on tap about each of us. They have access to more information than would be found in a run-of-the-mill MI5 file; the ease with which they can pry into the lives of others would make them the envy of the Stasi. The two surveillance corporates use this information to run vast, high-speed, computerised marketplaces that are opaque to outsiders and to market participants alike—and which are currently unregulated. What the companies have constructed are automated systems for enabling advertisers to target users with commercial messages. The commercial pitch was that such messages could be precisely calibrated to match the known predilections of the target. Advertisers’ acquiescence has led to a vast inflow of advertising dollars to Google and Facebook. But their systems have proved to be not only a licence to print money, but also a licence to destabilise democracy.
There is no disputing the reality of digital power, but it is often difficult to conceptualise using the intellectual frameworks of the analogue era. We find ourselves grappling with novel kinds of capability that call for novel forms of remedy.
Consider the so-called “Right to Be Forgotten” granted by the European Court of Justice (ECJ) in 2014, which gives European Union citizens the right to petition Google to have information about them removed from the company’s search results in Europe. To call it a “right to be forgotten” is not strictly accurate; it is merely a right to request that certain information not be listed in Google’s European search results—although in our networked world, this almost amounts to the same thing. If the dominant search engine doesn’t find you, then you have effectively ceased to exist. The ECJ ruling bows to the reality that Google has a unique capacity to make or break reputations.
Google itself has been given responsibility for managing the complaints and adjudicating who gets to be “forgotten,” effectively outsourcing a judicial responsibility to a private company. Territorial sovereignty, the kind exercised by elected governments, has been supplanted by what the legal scholar Frank Pasquale calls “functional sovereignty.” The digital giants, Pasquale maintains, “are no longer market participants.” Rather, “they are market makers” in their fields, “able to exert regulatory control over the terms on which others can sell goods and services.” Moreover, he says, “they aspire to displace more government roles over time, replacing the logic of territorial sovereignty with functional sovereignty. In functional arenas from room-letting to transportation to commerce, persons will be increasingly subject to corporate, rather than democratic, control.”
There is no issue so big that the tech companies think they can’t handle it themselves. When the issue of fake news flared up amid the US election, for example, the first response of Mark Zuckerberg, Facebook’s CEO, was a mixture of denial and incredulity. Then, as the evidence mounted that his advertising machine had been weaponised by dark political actors, he pivoted rapidly from that incredulity to scepticism and then—as the evidence became incontrovertible—to a technocratic determination to “solve” the problem. By the end of September, he was issuing a personal Yom Kippur post on Facebook pleading for “forgiveness” in light of the way that “my work was used to divide people.”
But Facebook has two conflicts of interest that inhibit it from fixing the problems. First, surveillance capitalism requires the maximisation of “user engagement,” to create the data that is to be monetised. And it turns out that Facebook users are often more engaged by fake news than they are by mundane truths. The much-vaunted, pending overhaul of Facebook’s algorithm to give priority to material shared between individuals represents a retreat from real news just as much as it does from fake. That could have adverse implications for the responsible media, and meanwhile, of course, that emphasis on shares only deepens the “engagement.” The second conflict stems from the fact that if Zuckerberg were to accept editorial responsibility for what is posted by his website’s users it would effectively destroy his company, given that there aren’t enough administrators in the world to vet what gets posted to Facebook in a single second.
This hasn’t stopped some from demanding that social media organisations accept responsibility for what appears on their sites. And the Germans have passed legislation that mandates swingeing penalties on platforms that do not take down offending content in a matter of hours. But like the right to be forgotten, this statute delegates to private companies the task of deciding what shall and shall not be published in a democracy—another illustration of functional sovereignty replacing its territorial counterpart.
But in Washington and London alike, the emerging digital giants were given a free pass for a couple of decades, irrespective of which party was in power. What’s more, they were lionised—and solemnly consulted at Davos. Before being elected in 2010, David Cameron allowed his party to float the idea of putting Google in charge of NHS records; his then policy adviser Steve Hilton is married to Rachel Whetstone, who worked for the search giant at the time. One can understand the reluctance to meddle: there were few easy answers, and politicians raised in the analogue era can be forgiven if they struggle to find what answers there are. These companies had, furthermore, grown so fast that their influence in day-to-day life only became apparent after they were already entrenched. The oldest of the five giants, Microsoft and Apple, have their origins in the mid-1970s, but Microsoft was of no significance until it landed a contract with IBM in the 1980s, and Apple’s resurgence only dates from Steve Jobs’s return in 1997. Amazon was only formed in 1994, Google in 1998 and Facebook in 2004. But whatever the reason, they enjoyed not so much a light as an absent touch where regulation was concerned.
This pattern of indulgence only started to change in 2017, as western governments began to realise the political clout that Google and Facebook might possess. The great fear is that their concern has come too late to do any good; they have reordered the world and, along the way, made addicts of its citizens.
Facebook’s aim is to capture the attention of its users—to distract them. Most of its efforts are devoted to creating addictive diversions which maximise that all-important (and monetisable) “user engagement”; in other words, providing services and products that offer instant gratification over reflection or even thought.
The wags who spoke of early smartphones as “Crack-Berry” (a play on BlackBerry, one early model) were not wrong; they are engines for dopamine release. In January, two major Apple shareholders urged the company to address the potential harms caused by its products. They advised the firm to increase the parental controls on its devices and research the mental health effects of excessive use of smartphones. The shareholders, with a combined holding of $2bn, wrote: “There is a growing body of evidence that, for at least some of the most frequent users, this may be having unintentional negative consequences.” A few days later, Roger McNamee, a large and early investor in Facebook, went public with very similar arguments. Indeed, the whole tendency for smartphones and social media apps is to “privilege our impulses over our intentions,” in the words of Oxford computer scientist James Williams. As a result, in Williams’s view, social media is a threat to liberal democracy, a system that must always rely on the ordinary citizen’s capacity for deliberation.
That may, perhaps, imply an unduly romantic notion of the voter’s capacity for political reflection. But that nagging sense persists that our capacity for citizenship, and perhaps our psychology more generally, is being warped. So conservative parents concerned about kids’ channels on YouTube suddenly find themselves agreeing with left-wing critiques of capitalist power. Sincere socialists and thoughtful libertarians who would normally be at loggerheads will, very likely, also find themselves in violent agreement when it comes to railing against the dominance of Facebook and Google.
The anxieties are newly pervasive, yet remain too shapeless to get much political traction. They do not come allied to any collective picture of where the companies may be taking us. A handful of artists and activists have tried to provide one: Dave Eggers’s 2013 novel The Circle, for instance, maps out the dystopian direction the Google/Facebook mindset is taking us in. But in general we’re suffering from a kind of imaginative failure.
Back in the 1980s, the cultural critic Neil Postman argued that our future would be bracketed by the nightmares of two British novelists: George Orwell, who thought we would be destroyed by the things we fear; and Aldous Huxley, who believed that our
undoing would be the things that delighted us. With the aid of digital technology, we are managing to achieve both nightmares at once. We click compulsively on health scares and other anxieties while Big Brother watches, but are also distracted by dubious political claims that make us feel good and reinforce our prejudices.
The traditional remedy for overweening corporate power is anti-trust legislation: that is, rules to prevent the formation of monopolies. But the legislation we currently have—especially in the US, which is where all the digital giants are based—is no longer fit for purpose in defining the monopoly problem.
Monopolies have always been deemed evil because they inflicted economic harm on customers. But how to define that harm today? In the distant past, anti-trust legislation used to hold excessive market share itself to be the problem. But Reaganomics narrowed the issue down to one of prices, holding back from interfering with mighty corporate giants if they could claim they were adequately disciplined by the mere potential for competition, so-called “contestability.” This post-1970s legal doctrine gives rise to paradoxes in the digital age. Since neither Google nor Facebook charges its users anything the harm isn’t immediately visible. In fact, being a monopoly in a digital marketplace might be a sign of superior quality and service. If so many people use Google when, say, Bing is just a click away, then Google must be doing something right, and to sue it for being a monopoly would be tantamount to punishing excellence.
Stuck in this frame of thinking, US regulators have been reluctant to pursue the digital giants. (Things are somewhat different in Europe: in 2017 the European Commission levied a whopping fine on Google for abusing its dominance.) The glaring flaw in the American argument is that Google’s (or Facebook’s) users are not its customers. It could well be that those who are—the companies who bid for advertising slots in the hidden, unregulated, high-speed auctions which both companies run—are the ones suffering real harm.
Any effective regulatory action must focus on these auction systems. Both companies have moved from being market participants to market makers, and an investigation into how they manage these auctions is long overdue. Advertisers themselves are now calling for one.
Another approach would be to revisit Section 230 of the Communications Decency Act (part of the sprawling US 1996 Telecommunications Act), which absolves the tech giants of legal responsibility for what passes through their servers. The key sentence reads: “No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.” Careful redrafting of this Section could—at a stroke—oblige social media companies to accept some degree of responsibility for what appears on their sites.
Political campaigning is yet another nettle to grasp. It has, of course, always been a dirty business, polluted by dark money and the unscrupulous use of whatever communications media are available at any given time. In one sense, 2016 merely confirmed that social media have become the contemporary tools of choice for manipulating public opinion. But they also have distinct advantages in the brainwashing game. For one thing, many of the campaign funds can be hidden from regulators. For another, bespoke and possibly contradictory messages from the same campaign can be sent to different individuals at the same time. Then there’s their cost-effectiveness. Investigations in the US suggest that Russian agencies spent just $30,000 on Facebook—less than one-hundredth of a cent for every US voter—and yet their messages reached up to 126m users.
Much tighter reporting of expenditure on data-analytics and social media campaigns during elections would help, as would a new requirement on Facebook and other social media companies to publish all the campaigning messages transmitted via their platforms.
In pondering what could be done to bring the new giants under democratic control, there’s a temptation to look backwards to earlier episodes in the history of capitalism when cycles of runaway growth led to overweening corporate power, followed by periods of reform. We think of Teddy Roosevelt and the Progressive movement of the late-19th and early-20th centuries which eventually led to the breakup of some huge industrial trusts—the railroad, oil and steel conglomerates that were then suffocating the US economy.
We are at a comparable point, facing five huge companies that own the technological platforms which will dominate social and economic life for the foreseeable future. Can we envisage an analogous contemporary movement to take back control? It seems unlikely. One of the reasons that Roosevelt was able to act so decisively was because there was significant—and focused—public concern about the activities of the “Robber Barons” of his day. They were crushing competitors and ripping off the public. Today, the anxieties about prying profiteers are just as real, but much vaguer, which makes it far harder to turn popular concern into a winning practical manifesto.
Furthermore, politicians will be wary of taking on industrial concerns with such deep pockets, and legions of addicted users. Or, at least, western politicians will be. In some undemocratic countries, notably China, the political will and commensurate resources have been found to bring the internet giants to heel, replacing them with state-sponsored alternatives—without impairing rapid growth of internet use. Much else that we would want to conserve, including freedom of speech, has been crushed in the process; but it is nonetheless a reminder that different political attitudes could see internet technologies run in very different ways. Still, it doesn’t seem wise for westerners to hold their breath hoping for anything comparably bold.
The language of addiction may seem strange when applied to people’s attachment to social media, but it seems increasingly apt. In conversation, one hears users complaining that they cannot understand how they spend much time on Twitter, Facebook, YouTube or Instagram, or why they find themselves unable to stop checking their smartphones. (One recent US study found that people touched their phones an average of 2,600 times a day.) Yet suggestions that they simply quit the habit are met with a grimace and a shrug. There’s a feeling of passive acceptance which is reminiscent of what one hears from people who believe that they might—just might—have an alcohol problem.
But if there is any validity in this analogy, then the implications are troubling. For one thing, liberal democracies mostly avoid interfering with the right of citizens to harm themselves and—up to now at least—the NHS does not turn away smokers with lung cancer or 35-year-olds with cirrhosis of the liver. There are no votes in temperance, it seems. And as the US learned in the 1920s, prohibition doesn’t work.
And yet one does not need to give up hope. Paradoxically, it is the ordinary users—all those addicts of whom I write—who hold the key to a better future. For the one thing which truly terrifies Google and Facebook is that people might stop using their services. If one million people decided to take a break from Facebook for a month, say, then that alone would have a dramatic impact on the company, undermining its narrative of continual expansion and growth. More importantly, a million people would suddenly find that it’s possible to kick the social media habit. And they’d have an extra hour in their day to do something really interesting.