YouTube has consumed a good part of my days for more than a decade. As a teen-ager, I used the video-streaming platform to scrounge for crumbs of knowledge, watching free lectures on everything from algebra to literary modernism. Now I navigate to the YouTube app on my television most mornings to watch the news. I stream workout videos. I listen to music. I watch celebrities give tours of their garishly decorated mansions. Sometimes I stay on the site for hours, lost in the maze of memes, dinner ideas, and all manner of distraction.
My YouTube habit is far from unique. According to the company, the site has more than two billion monthly “logged-in” users. In a given twenty-four-hour period, more than a billion hours of video are streamed, and every minute around five hundred hours of video are uploaded. The torrent of content added to the site has helped establish new forms of entertainment (unboxing videos) and revolutionized existing ones (the mukbang). YouTube is a social network, but it is more than that; it is a library, a music-streaming platform, and a babysitting service. The site hosts the world’s largest collection of instructional videos. If you want to fix a tractor or snake a drain or perfectly dice an onion, you can learn how to do these things on YouTube. Of course, these are not the only things you can learn. Anti-vaxxers, 9/11 truthers, live-streamed acts of mass violence—all of these have surfaced on YouTube, too.
“No company has done more to create the online attention economy we’re all living in today,” Mark Bergen writes at the start of “Like, Comment, Subscribe,” his detailed history of YouTube, from 2005, the year it was founded, to the present. Among the titans of social media, YouTube is sometimes overlooked. It has not attracted as much adulation, censure, theorizing, or scrutiny as its rivals Facebook and Twitter. Its founders are not public figures on the order of Mark Zuckerberg or Jack Dorsey. Aaron Sorkin hasn’t scripted a movie about YouTube. But Bergen argues that YouTube “set the stage for modern social media, making decisions throughout its history that shaped how attention, money, ideology, and everything else worked online.” It’s one thing to attract attention on the Internet; it’s another thing to turn attention into money, and this is where YouTube has excelled. The site, Bergen writes, was “paying people to make videos when Facebook was still a site for dorm-room flirting, when Twitter was a techie fad, and a decade before TikTok existed.” Posting on Facebook or Twitter might net you social capital, an audience, or even a branded-content deal, but the benefits of uploading videos to YouTube are more tangible: its users can get a cut of the company’s revenue.
The site has been compensating “creators” since 2007, a scant two years after it launched, and only a year after Google acquired the company for a price tag of $1.65 billion. YouTube splits its advertising revenue fifty-five per cent to forty-five per cent, in favor of creators—one of the best deals available to anyone hoping to be paid for their time on the Internet. Since 2018, the main prerequisites a creator has needed to monetize their videos is a minimum of a thousand subscribers and four thousand “watch hours” in the previous twelve months. Recipe developers, video-game live-streamers, podcasters, teen-age trolls, children playing with toys, aspiring entrepreneurs hawking get-rich-quick schemes, right-wing shock jocks (at least those who haven’t been demonetized), and major television networks are all members of the baronial class of YouTube moneymakers. In a recent interview, the veteran science-and-education vlogger Hank Green said that the site presented such favorable terms that the idea he would “walk away” from YouTube would be like leaving America: “There are things I very much do not like about it, but I feel a little like a citizen, so that would be such a big decision to make.”
Bergen, a reporter for Bloomberg News and Businessweek, catalogues YouTube’s rise and the billions (of users, dollars, hours of video) it controls in a tone that is at once resigned, rhapsodic, and disgusted. The story his book unspools is one of breathtaking profit and foolish stumbles, violence and greed and corporate obfuscation. It is also one of surprising stability: YouTube, Bergen writes, is “the sleeping giant of social media.” Even as TikTok has become a megalith and other social networks have lost their touch with the youth, the site has retained its audience. A recent Pew Poll found that YouTube is used by ninety-five per cent of American teen-agers aged thirteen to seventeen, compared to sixty-seven per cent who used TikTok. As one of its employees told Bergen, “How do you boycott electricity?”
YouTube was the invention of three former PayPal employees: a graphic designer named Chad Hurley and two coders, Steve Chen and Jawed Karim. Hurley had the vaguely populist aspiration of providing a service for what he called the “everyday people” of Web 2.0, those online diarists who flocked to LiveJournal and WordPress. Inspired by early social-media sites such as Friendster and hornier fare like the attractiveness-ranking site Hot or Not, the trio hoped to create something fun, popular, and possibly even sexy. Dating was, in fact, tossed around as one of the early motivations for the site: in a memo Karim sent to his co-founders, he wrote, “A dating-focused video site will draw much more attention than stupid videos. Why? Because dating and finding girls is what most people who are not married are primarily occupied with.” (This hormone-fuelled premise, of course, informs the origins of Facebook, too.)
One of the book’s insights is that there’s no way to separate the economic power of YouTube from its emotional and psychological attachments—voyeurism is what inspired it in the first place. From this perspective, Hurley’s desire to cater to the “everyday people” of the Internet might be better understood as a prescient sense that one could convert the passive desire to watch and be watched into profit. As it has grown, YouTube has embraced this understanding of its purpose. A 2017 video outlining the company’s brand mission claims that the site’s purpose is to discover “the rawest, purest, most unfiltered portrait of who we are as people.” The drive to make ever more attention-getting content has created a persistently difficult question for the site: What are you willing to do to capture eyeballs?
Any platform that relies on user-generated content is always at war with its more troublesome users—those uploading not only pirated stuff but also material of the shocking and morally reprehensible variety. Content moderators were an essential part of YouTube’s early staff. The challenge they faced wasn’t just nudity or content that broke copyright rules but an unceasing geyser of grossness—“fetish clips of women in heels stepping on creatures, cats being boiled alive”—and images so disturbing that one former moderator would only barely describe to Bergen what she saw. Other content violations were more subtle. An early handbook implored moderation staff to “Use your judgment!” above a pair of pictures. One that showed, according to Bergen, a “woman holding a banana to her mouth suggestively” should be rejected, though “another eating a corn dog normally” was “fine.” One of the lawyers who helped outline YouTube policy in those days wondered, “What kind of Pandora’s box have we opened?” Bergen deploys his own mythic comparison to the site’s unruly, babbling collective of content: “The sprawling, Babelian video site made internet governance nearly impossible.”
As early as 2007, mass murderers were taking to YouTube to air out their beliefs and share their plans. The site’s moderators and other warning systems (concerned viewers could flag videos, which then entered a lengthy queue) could not always respond quickly. A year earlier, the Thai government threatened to ban the site for hosting videos that insulted the king, a criminal offense in the country, leading a Google lawyer to urge a YouTube contractor to leave as quickly as possible. Around the same time, German officials bombarded the company with demands to remove videos containing Nazi imagery. Since YouTube did not yet have an office in Germany, the company decided that it could get away with ignoring the requests; a member of the policy team placed a placard on his desk that read “Do not Appease the Germans.” As the mountains of content piled up, and YouTube’s scope became more global, one thing became abundantly clear: imploring someone to use their judgment was a flimsy model for best practices.
The company used Google’s technology to develop artificial-intelligence tools to scan footage for clear content violations, allowing for the removal of obvious hate symbols, sexually explicit content, and copied footage. It struggled to articulate a consistent rationale for other decisions—offending content was handled in ways that could seem slipshod, panicked, and self-serving. YouTube removed some videos deemed offensive (for example, the more violent variety of Saddam Hussein-execution clips) and left others up (a teaser for an Islamophobic movie that led to protests around the Middle East). When criticized, the company turned to what has become a standard defense among the social-media giants: it was merely a platform and not a publisher. But YouTube, Bergen’s book makes clear, was never a neutral arbiter—the company made decisions that influenced which ideas succeeded. A series of changes to its recommendation algorithm, beginning in 2012, illustrated a clear preference for certain kinds of content: “watch time” was privileged over “views,” meaning that videos that kept viewers engaged for longer were given preferential treatment, and “viral hits” that attempted to achieve views alone were downgraded.
The algorithm changes, or at least the perception of them, affected the kind of videos that were posted to the site. Out were ephemeral joys, such as clips of dogs on skateboards. In were videos that the recommendation engine seemed more likely to surface: “long, engaging content about something newsworthy,” according to Bergen. This included videos by unsavory talking heads, such as the Canadian white nationalist Stefan Molyneux and the British men’s-rights activist known as Sargon of Akkad, who dilated on race science and anti-feminism—subjects that attracted a particularly devoted core of viewers. But the single greatest beneficiary of the algorithm shift was the Swedish video-gamer Felix Kjellberg, known as PewDiePie. His hours-long live streams and antic commentary were perfectly suited to an algorithm that rewarded eyeballs on the screen. By the early twenty-tens, he had become YouTube’s biggest star. As his persona grew more extreme, he became one of the company’s biggest problems, attracting constant attention for provocative stunts. YouTube apparently put up with him for a simple reason—he was a cash cow. From 2012 to 2019, Bergen writes, “humanity consumed 130,322,387,624 minutes of PewDiePie videos.” The ads shown during these videos earned him more than thirty-six million dollars and YouTube itself around thirty-two million dollars.
For a number of years, YouTube largely left alone more expliclity bigoted and conspiratorial figures such as Molyneux, Richard Spencer, and Alex Jones. Most of them, as Bergen writes, “were also on Facebook and Twitter anyway.” A 2017 investigation from the Times of London found that videos uploaded by neo-Nazis and ISIS supporters were benefitting from ad placement; companies like A. T. & T. and Johnson & Johnson threatened to boycott. YouTube, along with the other social-media titans, began to ban the likes of Jones and Molyneux, and demonetized the borderline cases as it revised and revised again its “hate speech” policies. But a cottage industry of take artists continues to flourish. This spring, creators commenting on (or mocking) the Amber Heard–Johnny Depp defamation trial were popular and profitable. (Though some recent research suggests that the algorithm may not have bolstered YouTube’s fringe as much as the platform’s basics of mass scale and economy did.)
Yet, despite a continual stream of scandal and disaster, YouTube has emerged from its controversies relatively unscathed. The site is not immune to public pressure or the threat of regulation—Google and YouTube paid a hundred and seventy million dollars in a settlement with the F.T.C. and the State of New York over allegations that it had violated children’s online-privacy laws. (This month, the Supreme Court agreed to hear a potentially momentous case: the outcome of Gonzalez v. Google will decide if YouTube is more broadly liable for the content it makes available for its user.) But YouTube doesn’t have the outsized reputation that Facebook and Twitter have taken on as public villains, destroyers of democracy, and general irritants. How has it evaded the same characterization?
Bergen comes up with several answers to this question. He suggests that, because YouTube’s most popular videos are music, video-game streaming, and children’s entertainment, its political content has attracted less attention than Facebook’s. He also notes that the platform was “better situated to avoid information warfare” than Facebook or Twitter because the company shared “relatively little data with outsiders,” and therefore played a less prominent role in debates over misinformation. Crucially, Donald Trump was not popular or active on YouTube either, and that allowed the site to avoid the scrutiny of the other platforms that served as his bully pulpit. But a more convincing reason for why YouTube has avoided becoming a public punching bag like its declining rival Facebook is that it is simply too useful and too ubiquitous to fail. As long as the platform is where everyone on the Internet goes to do their homework and fix their plumbing, it will continue to be one of the Internet’s indispensable sites—a self-sustaining ecosystem that incentivizes the creation of more and more content. What this content is doing to us, its viewers, is another question. As a YouTube employee put it in a personal newsletter at the end of 2014, “With over 72 billion hours of video watched on YouTube this year—mostly torpid fragments of pop culture—it IS possible we’re seeing too much.” ♦
