Antisocial, page 9
A few nerdy young men, most of whom accepted the basic tenets of techno-utopianism by default, created early versions of blogging software: LiveJournal, WordPress, Blogger, Movable Type. Soon afterward, another cohort of nerdy young men founded a few fast-growing social networks: MySpace, Reddit, Twitter, Facebook. They didn’t pretend to know exactly how social media would be used, and they gave even less thought to how it might be misused. They wanted to “change the world,” but they didn’t bother specifying that they wanted to change it for the better—that part was implied, and besides, it was supposed to happen more or less automatically.
Even after they started to accrue vast wealth, they kept referring to themselves as disrupters, or hackers. They wielded unprecedented power, but they seemed uneasy acknowledging it, much less deciding what to do with it. They often implied that their amplification of all voices, even the most corrosive, was an inevitable by-product of technological progress. In fact, it was a choice.
* * *
—
In early 2012, Facebook announced its intention to become a public company. On page 67 of an SEC filing, right after sections about inflation risk and interest-rate sensitivity, was an open letter signed by Zuckerberg, Facebook’s twenty-seven-year-old CEO, soon to become its majority stockholder. “At Facebook, we’re inspired by technologies that have revolutionized how people spread and consume information,” he wrote. The letter claimed that Zuckerberg and his employees were wont to stand around their open-plan office, chatting about their eagerness to democratize global discourse. “We often talk about inventions like the printing press and the television—by simply making communication more efficient, they led to a complete transformation of many important parts of society,” he wrote. “They encouraged progress. They changed the way society was organized. They brought us closer together.” This story wasn’t entirely wrong, but it left out a lot.
On the day Facebook went public, Zuckerberg’s personal fortune increased by more than $8 billion. Still, he kept up a strict semiotic regimen that downplayed his power. He worked at a bullpen desk, alongside a few of his two thousand employees, in the company’s headquarters in Menlo Park (vanity address: 1 Hacker Way). He wore a daily uniform of a gray T-shirt and jeans. (The T-shirts cost about three hundred dollars each, but he didn’t publicize that fact.)
It was a boom time for techno-libertarianism. Barack Obama, whose political ascent had been facilitated by an online groundswell, often spoke in a tone of optimism, even utopianism, about the salubrious effects of social media. “In the twenty-first century, information is power,” he said in a 2011 speech on Middle East policy. “The truth cannot be hidden.” Social media entrepreneurs, eager to be perceived as dispassionate and democratic, pledged to keep their platforms “content neutral.” If they thought of themselves as gatekeepers at all, they seemed to take for granted that the gates ought to be thrown wide open. Twitter’s executives often referred to their company as “the free-speech wing of the free-speech party.” This all seemed, at first glance, like an unambiguous victory for freedom, and who didn’t love freedom?
And yet, the more closely you looked, the less obvious it all seemed. The First Amendment applied only to the government, not to private businesses. Instead of citing the First Amendment itself, then, social media companies invoked fealty to analogous “free-speech principles.” Which principles, exactly? For centuries, the meaning of free speech had been refined and reinterpreted in universities, in legislatures, in the courts, in the press.* In Silicon Valley, however, weighty decisions about free speech might be made in the course of an afternoon, in a cramped conference room full of complimentary bottles of seltzer and kombucha, by a small team of harried computer engineers. Often, they had no long-term plan other than hacking together a “minimum viable product,” “shipping” their code as quickly as possible, and then “iterating”—all start-up euphemisms, essentially, for trial and error.
The disrupters had gleaned, through cultural osmosis, that free speech was a value worth protecting. Beyond that, they weren’t expected to spend much time thinking through the underlying principles. Instead, they released their products into the world and then waited to see what would happen. In their most optimistic moments, they could convince themselves, and sometimes everyone else, that the internet had finally eliminated the problem of gatekeepers. But it had only shifted the problem.
At a public event in Rome in 2016, a few hours after a private audience with the pope, Zuckerberg was asked whether he saw himself as an editor. “No,” he said, tittering uncomfortably. “We’re a technology company, not a media company. . . . We build tools. We do not produce content.” In other public settings, he tested out slight variations on this argument. Sometimes he tried to absolve himself of decision-making power; sometimes he acknowledged his power, but framed his actions as compulsory, or inherently noble, implying that the freedom to share opinions online was akin to a human right. Sometimes he deployed several dodges, one after another, in the tradition of William Caxton: information wants to be free; besides, people who take offense at what they’ve read should blame the author, not the messenger; anyway, the ultimate responsibility lies with each individual reader.
Zuckerberg repeatedly insisted that Facebook was a platform, not a publisher. If some disgruntled teenager wanted to quote Socrates’ vituperative opinions about women—or if, for that matter, a teenager wanted to share his own vituperative opinions—then who was Zuckerberg to stand in the way? He might not personally endorse every view expressed on his platform, but he believed in giving power to the people.
Each individual social media feed became a unique and unpredictable blend of fact, satire, rumor, propaganda, alarmist diet tips, and advertainment. There were not enough Caxtons to go around. For every vile or propagandistic post that got weeded out, a hundred others bloomed. In the United States, the disruptive effects of social media coincided with a period of stark economic inequality, cultural unrest, and rapid demographic change. In 2013, for the first time in American history, a majority of infants in the country were nonwhite, a fact that many white Americans perceived, consciously or unconsciously, as a threat. This was the kind of topic that was difficult to discuss productively even under ideal conversational conditions. The conditions on social media, to put it mildly, were far from ideal.
Still, the new internet platforms continued to flourish, empowering both prosocial and antisocial voices. One essay on Return of Kings, a well-known “neomasculinist” blog, cited Plato, Aquinas, and Aristotle. The headline was “Mate, Hate is Great! A Philosophical Defense of Misogyny.”* Another blogger, writing under the pseudonym Quintus Curtius, envisaged “a future where classical knowledge will be driven underground . . . as not being in tune with modern feminism and political correctness.” He warned his readers that “the commissars of modern culture don’t want you to know too much about history, or about how things were like in previous eras.” When he wrote these words, in 2014, the most powerful informational gatekeepers in the country included some of its biggest tech companies: Amazon, Twitter, Facebook, and Google. And yet Quintus’s dystopian reverie was published on a proudly misogynist blog that enjoyed a verified Twitter account, a popular YouTube channel, a podcast on iTunes, and a prominent position in Google’s search ranking. His self-published book of “essays on life, wisdom, and masculinity” was selling briskly on Amazon. The “commissars of modern culture” were not censoring Quintus’s noxious brand of male supremacy. If anything, they were promoting it.
PART TWO
A Human Superpower
The internet is almost the perfect distillation of the American capitalist ethos, a flood of seductive choices. . . . I can’t think of a better summing up of what America’s strengths and weaknesses are right now.
David Foster Wallace, 2000
CHAPTER FIVE
The Gleaming Vehicle
In April 2014, looking for new story ideas, I attended a tech conference in a stylish hotel in Lower Manhattan. The conference was called F.ounders, a word that no one, including the founders of F.ounders, could decide how to pronounce. Half of us stammered over the stray period. The other half ignored it. It stood for nothing, apparently, except for the general concept of innovation.
I arrived just in time for cocktail hour, which was being held in a well-appointed room with sweeping sunset views of the lower Hudson. An organizer handed me a complimentary tote bag, a laminated badge bearing my name and job title, and a little black book—one of the few paper facebooks I’d seen since the advent of Facebook—containing the headshot and bio of every entrepreneur, engineer, and venture capitalist in attendance. I flipped through it, looking for names I recognized.
One was Eli Pariser, an activist turned entrepreneur. Pariser was the former director of MoveOn.org, a left-wing nonprofit specializing in online organizing. More recently, he’d written The Filter Bubble, a book demonstrating that as huge tech companies grew huger, they were fracturing the Internet into millions of bespoke internets. The more personal data they had about you, Pariser warned, the better their algorithms would get at showing you links you’d feel compelled to click on. This was called microtargeting. To the tech companies—and to advertisers, propagandists, or anyone else with a message to spread—this seemed like a win-win. But for American democracy, Pariser argued, the consequences could be dire. “Most personalized filters have no way of prioritizing what really matters but gets fewer clicks,” he wrote. “In the end, ‘Give the people what they want’ is a brittle and shallow civic philosophy.”
At the time, Google owned almost 40 percent of the online advertising market, and Facebook owned another 10 percent. Some analysts were already warning that they might comprise a duopoly. Both companies’ business models, especially Facebook’s, were built around microtargeting. Filter bubbles, in other words, were not a temporary bug but a central feature of social media. It was hard to see how the latter could flourish without the former. If filter bubbles were bad for democracy, then, were Google and Facebook also bad for democracy?
It was a fair question, almost an obvious one, and yet the cultural vocabulary of the time did not allow most people to hold it in their heads for long. Pariser’s book was published in 2011, the year of the Arab Spring—which was organized, in part, via social media, and which was often called the Twitter Revolution. Mark Zuckerberg had just been named Time’s Person of the Year; in the hagiographic cover photo, his eyes were oceanic and farseeing, dreaming up ingenious new ways to forge human bonds. If some movies and books portrayed him as shifty, even a bit ruthless, it was still possible to imagine that ruthlessness, in the tradition of Thomas Edison or Steve Jobs, was merely the cost of doing business.* Zuckerberg’s motto, “Move Fast and Break Things,” was generally treated as a sign of youthful insouciance, not of galling rapacity. Facebook’s users—more than a billion of them—seemed happy. Its investors were delighted. If social media wasn’t a good product, then why was it so successful?
At the time, it was still considered divisive (at swanky New York tech conferences, anyway) to wonder whether the behoodied young innovators of Silicon Valley might turn out to be robber barons. It was far more socially acceptable to extol the gleaming vehicle of technology—to gaze in amoral awe at its speed and vigor—than to ask precisely where it was headed, or whether it might one day hurtle off a cliff.* Such questions had come to seem fusty and antidemocratic; people who spent too much time worrying about them were often dismissed as cranks or Luddites. To a techno-optimist, there was only one way the vehicle could possibly be going: forward.
So, in Silicon Valley and in much of the national press, The Filter Bubble was taken as mild, constructive criticism.* Pariser was treated not as a lone protester staring down a phalanx of tanks but as a decorous bystander asking the tank commanders to turn their wheels a few degrees to the left. Some of the new gatekeepers denied the existence of filter bubbles; others acknowledged the problem, privately or even publicly, but seemed either unwilling or unable to do much about it. In the meantime, Pariser reasoned, if he couldn’t change the way the content on social media was filtered, maybe he could change the content itself. So, the year after his book came out, Pariser went into the content-aggregation business. He cofounded Upworthy, a company whose posts were designed to go viral on Facebook.
* * *
—
It wasn’t news to anyone at F.ounders that the traditional news industry was collapsing. Some of the fine points were debatable, but the contours of the recent history were clear enough. First, around the turn of the century, came the transition from print to web. Instead of buying a newspaper at a physical kiosk, now you could browse an infinite online kiosk for free. Instead of paying to run a classified ad, you could post one, in seconds, on Craigslist. The internet had released a cascade of information: countless international wire services, ancient scriptures, legal libraries, and anime message boards, all available instantaneously. For newspaper publishers, this was a devastating threat to their business models. But for many readers, although the web made their lives more convenient, it didn’t change their reading habits very much. If you were used to scanning the front page of The Denver Post every morning, you could now scan the Post’s home page instead.
By 2014, the industry was in the midst of a more profound shift: the transition from the open web to the social web. You could still start your day by going straight to denverpost.com, but, statistically speaking, nobody did that anymore. Now you went to Facebook or Google or Reddit or Twitter, where all the world’s information sources—The Denver Post, The Denver Guardian, your ex-pastor, your estranged aunt, Alex Jones, Van Jones, Geico, a twelve-year-old influencer from Norway—clamored for your attention within a single stream. In the TV era, gatekeepers had controlled the flow of information, but at least the consumer could always change the channel. In the Facebook era, the browsing experience felt so passive, so close to nonvolitional, that the standard metaphor was no longer consumption but viral infection.
Denver Post reporters knew how to cover a mayoral race or a Rockies game. Most of them didn’t know, and didn’t care to learn, how to promote their articles on social media. Upworthy, by contrast, was built around the dual goals of “clickability” and “sharability.” It was still true, as Pariser had pointed out in his book, that social media algorithms had “no way of prioritizing what really matters.” But now, instead of advocating for the algorithms to be fixed, he hoped to adapt to them, manipulating Facebook’s insentient censors into letting a few chosen links go viral.
At the time, Upworthy didn’t create any content. Instead, it scoured the web for short videos about hope and human triumph, then repackaged those videos with catchier headlines, more tantalizing thumbnail images, and the like. Each change was intended to make a post slightly more clickable and sharable, the way each tweak to a race car’s body made it a bit more aerodynamic. Old-school journalists were trained to spend most of their time thinking about reporting, writing, and fact-checking; the headline was generally an afterthought. Upworthy, responding rationally to the upside-down incentives of the social web, outsourced content production entirely and focused instead on headline writing. At least twenty-five headlines would be generated for each post; a few rival headlines would then be tested against one another, algorithmically, to determine which was best. (It went without saying that “best” was synonymous with “most popular.”)*
By the spring of 2014, Upworthy and its many imitators had developed a formula for making headlines as clickable as possible. The headlines often conveyed just enough information to be enticing but not quite satisfying, leaving a “curiosity gap.” A Facebook user scrolling through her feed could only satisfy her curiosity by clicking on the link, giving Upworthy another pageview. (A Montana slam poet’s video about self-love, which she’d posted to YouTube as “Self love poem,” became, on Upworthy, “She Lied to Herself Every Night for a Year. Here’s What Happened When She Was Honest.”) This headline style came to be known as clickbait. It was easy to parrot, or to parody; mainstream outlets did both, winking at the phenomenon while profiting from it. The Onion launched a spinoff site, ClickHole, devoted to satirizing the new genre.* There were also “clickbait spoilers,” such as the Twitter account @SavedYouAClick, whose only purpose was to spoil Upworthy-style headlines (“Julia Roberts Will Literally Never Have a Bigger Role”) by providing the missing information (“The voice of Mother Nature in a commercial”). And yet, despite all this, clickbait worked astoundingly well. A year after it launched, Upworthy reported 87 million unique visits a month—more than The New York Times.
Curiosity is not the only way to get clicks, of course. Humor also works, as do lust, and nostalgia, and envy, and outrage. There are as many ways to attract a person’s attention as there are to bait a mousetrap, and some baits work better than others. “Content that evokes high-arousal emotion is more likely to be shared,” two Wharton professors wrote in 2012. “Positive and negative emotions characterized by high arousal (i.e., awe, anxiety, and anger) are positively linked to virality, while emotions characterized by low arousal (i.e., sadness) were negatively linked to virality.”
High-arousal emotions are also called activating emotions. They are emotions that lead to measurable behaviors—in this case, clicking or liking or sharing a link—as opposed to deactivating emotions, which are more likely to induce torpor or paralysis. In real life, of course, both kinds of emotion have their place. When a well-adjusted person faces a setback—say, a death in the family—a deactivating emotion is often the only appropriate response. On the viral internet, however, deactivating emotions are merely market inefficiencies. From the standpoint of sheer entrepreneurial competition, what matters is not whether a piece of online content is true or false, responsible or reckless, prosocial or antisocial. All that matters is how many activating emotions it can provoke.*
