The Beast With A Billion Eyes

By Lev Grossman

For every minute that passes in real time, 60 hours of video are uploaded to YouTube.

You can turn that number over in your mind as much as you want; at no point will it stop being incredible. Sixty hours every minute. That's five months of video every hour. That's 10 years of video every day. More video is uploaded to YouTube every month than has been broadcast by the three big TV networks in the past 60 years. And the pace is accelerating: last year the rate was only 48 hours per minute. William Blake once wrote something about seeing a world in a grain of sand and holding eternity in an hour. YouTube hasn't reached that point yet, but it's well on its way.

There's never been an object like YouTube in human history. It gets 4,000,000,000 page views a day, which adds up to 1,000,000,000,000--that's a trillion--a year. It has 800,000,000 users (about the same as Facebook) who watch 3,000,000,000 hours of video a month (that's 340,000 years). Human civilization now generates massive quantities of video footage simply as a by-product of its daily functioning, much as some industrial processes generate toxic slurry. Before YouTube there was no central catchment for all that video; now it drains into a single reservoir, where we as a species can pan through it and wallow in it endlessly.

In October 2006, when Google bought YouTube for $1.65 billion, it seemed possible that the company had screwed up on an epic scale--YouTube's rivals would split the market, or the Net would crash from all that video traffic, or YouTube would be sued into nothingness over copyright violations. But no: YouTube's rivals failed to thrive, the Net has held, and YouTube hacked out elaborate technological solutions to its copyright woes that even the plaintiffs had to admit were pretty cool. YouTube has eaten everything in sight.

YouTube's morbid obesity is a mixed blessing. It's a good thing for us in that it's handy to have all that video in the same place. It's good for Google, because check out all that traffic. But YouTube is becoming a difficult proposition for Google. After all, YouTube isn't like television. When you turn on a TV, you're presented with a limited number of options, which you know something about and which you can count on to be fairly professional-looking. On YouTube, the search engine is sifting through a billion options, literally, and you hardly know anything about any of them. You can't just turn YouTube on and chill out the way you do with TV.

This accounts for the one very small number among YouTube's many giant ones: 15 minutes. That's how long the average user spends on the site per day. Whereas the average American spends nearly three hours a day watching TV. And make no mistake: TV is the competition here.

The other consequence of YouTube's runaway success is that it's expensive: it costs Google a lot of money to keep the billion-eyed beast alive. It has to keep a lot of servers humming to store all that video, because YouTube never forgets, and it needs big, fat expensive pipes to keep those videos streaming 24/7, 365. Google isn't made of money.

Well, maybe it is. I'll get back to you on that. But even Google has to make YouTube pay, by running ads next to or before or on top of all those videos. Whatever else it might be or want to be, YouTube must be a gigantic, billion-faced advertising billboard. It has to consume attention and excrete cash.

Obviously, YouTube has no problem getting attention; it's just not necessarily the kind of attention advertisers like. All kinds of weird, random stuff gets uploaded to YouTube, and then other weird, random stuff gets appended to that stuff as comments, and advertisers don't like that. They don't want to get weird, random stuff all over their nice, clean brands. As a result, Google can't charge as much for its ads as, say, Fox does. It can't even charge as much as Hulu.com which gets a fraction of YouTube's traffic but which shows only professionally produced and sanitized content.

But Google can't just clean up YouTube, whack it on the ass and point it in the direction the company wants it to go. That would destroy YouTube's essential, anarchic nature, and anarchy is the beating heart of the billion-eyed beast. What to do?

In biology there's something called the square-cube law, which explains why animals can't stand up when they get too big. Likewise in astrophysics, even light is unable to escape black holes because they are so dense. An analogous fate constantly stalks YouTube, just over the event horizon: its messy, ungovernable enormousness is both its great strength and its fatal flaw. Which is why, while you and I are blowing off work to watch a monkey ride backward on a pig, a small army is fighting a ceaseless, Sisyphean war to keep YouTube from collapsing under its own weight. If YouTube can actually win that war, its victory will have consequences for the entire universe of broadcast media.

The headquarters for that army is in San Bruno, Calif., just south of San Francisco. (A hill near YouTube's office, one of those velvety California foothills that look like giant paws, bears a Hollywood-style sign that reads SOUTH SAN FRANCISCO, THE INDUSTRIAL CITY.) YouTube is housed in a huge, airy blond-wood office building; it appears to be part aerodrome but is in fact a former showroom for the Gap. (On the day I visited, YouTubers were rubbernecking at the sight of someone's office toy, a shark-shaped inflatable dirigible, that had gotten sucked into an air vent.) The centerpiece of YouTube's offices is a big, wavy red plastic slide in one of the atriums that runs from the top floor all the way to the bottom. You can find, somewhere on YouTube, a video of a dude catching big-time air and then totally wiping out on that slide; the legal disclaimer at the top of the slide is long and artful. Conference rooms are named after viral videos: you can take a meeting in Sad Keanu, or This Is Sparta!, or Socially Awkward Penguin. I spent the day in Nyan Cat.

(In case you missed it, Nyan Cat is one of the more mysterious and quintessentially YouTubey video phenomena ever to have crossed the viral threshold. It's a video of a cartoon cat with the body of a Pop-Tart that runs--scampers, really--through outer space, with a rainbow trailing behind it. Nyan Cat was originally created as an animated GIF by a 25-year-old Texas cartoonist named Chris Torres, who posted it on a comics website on April 2, 2011. Three days later, a YouTube user who goes by sarajoon [her bio reads, "I am the video word made flesh"] added a sound track, a maddeningly cheery Japanese pop song called "Nyanyanyanyanyanyanya," and reposted the GIF to YouTube as a video. The mashup went viral and finished 2011 as the fifth most watched video on YouTube, with 58,949,289 views. Wait--make that 58,949,290.)

The first line of defense against YouTube's runaway chaos, and the first angle of attack for a hopeful viewer, is search. YouTube gets a billion search queries a day; if they were tallied separately from Google's, YouTube would be the second largest search engine on the Internet. But searching for videos isn't like searching for Web pages. It's harder. Computers can read Web pages because they are made out of words: if you're looking for information on echidnas, a search engine can send you to a website where the string of letters "echidna" occurs a lot. But a computer can't watch a movie. It can't look at the huge string of 1's and 0's that make up a video file and know that that movie depicts a spiny, insectivorous, egg-laying mammal native to Australia and New Guinea. Computers are blind and deaf. To a computer, a movie is a black box.

This is one of YouTube's core organizational challenges. We help guide the poor, blind, deaf computers by attaching verbal descriptions to our videos, but unfortunately, we're not very reliable. Sometimes we'll title a video something like "For your viewing pleasure, this is a short film of a monkey riding backward on an echidna," but we're just as likely to call it "LOLOLOLOL this thing is amazeballs!!!!!!!" We're unpredictable that way.

YouTube can't watch videos, and it can't trust what we say, so instead it watches what we do. If you search for "echidna," YouTube will notice which of the search results you click on and will infer that that video is more echidna-y than the others; next time, it will be ranked a little higher. It will also notice if you watch the whole video or give up in the middle, and which video you watch right after it, and whether you post that video on your blog, and if you leave the site after you watch it or hang around for a while. It uses all that information to deduce things about the contents of the videos and improve its search results accordingly.

But it's not a permanent solution. Even as YouTube builds up its hoard of user-behavior data, the Youniverse keeps expanding. Moreover, search is useful only for people who know what they're looking for, and more often than not, people don't. Users tend to arrive at YouTube's front door the same way they sit down in front of a TV: with what is known inside the aerodrome as "low intent." They have no plan. They know they want to be entertained, but how or with what, they have no clue. And you can't search if you don't know what you're searching for; in fact, top search terms on YouTube include such plaintively vague requests as "funny videos" and "lol."

If YouTube is going to survive, and make money, and circumvent the square-cube law, new tactics will be required. With that in mind, the people who run YouTube are completely rebuilding it. YouTube needs to pull itself together and help users elevate their intent. If it can do that, it can start to compete with TV--and maybe even, as the beast has been known to do to its rivals, devour it.

You probably haven't thought much about YouTube's background color. Fortunately somebody is thinking about it for you: a smart, intense woman named Margaret Gould Stewart, whose business card reads director of user experience. Stewart oversaw the redesign of YouTube that began to roll out in December and continues in the form of ongoing changes. "We're never done," she says. "Literally, week to week, we're always tweaking."

The challenge Stewart faces is to create a container--her word--that will fit all of YouTube's vastly diverse content and make the task of navigating that content easier. That's why YouTube's background color, which used to be white, is now gray. "When you mat photographs," she says, "quite often you use gray instead of white or black because it tends to bring out a lot more of the nuances in a photograph."

Stewart's team also adjusted the height of YouTube's buttons and the radii of their rounded corners, and changed the way links look (they used to be blue and underlined; now they're just gray). "When you have it all blue and underlined in the default resting state, it really distracts," Stewart says. "This allows people to access that information when it's relevant, but it doesn't shout at them the whole time." Her team also enlarged the thumbnail images very slightly. "That change alone increased clicks to the Watch page by 2%," she says. "We were pretty amazed. We knew it was going to impact user behavior. We just didn't know how."

These are cosmetic changes, obviously. More changes are afoot, of a deeper, structural kind, changes in the very bones of the Youniverse. Whereas the old front page of YouTube was a grid of videos arrayed hopefully in front of you, chosen by some invisible hand for inscrutable reasons, now you get something much more civilized. The first things you see are a tidy list of channels on the left-hand side and a Facebook-style feed running down the middle of the page, consisting of recent videos from those channels on the left.

Broadly speaking, the channels are places where a bunch of videos can be grouped together onto one page by an individual user. They've always been there, but they used to be anemic things that most users ignored. "Users on YouTube really understand one noun, what we call a core object, and that's the video," says Noam Lovinsky, a group product manager at YouTube. "They don't really understand the concept of channels." YouTube is going to teach them. Where there used to be two units of organization on YouTube--a single video and the 1 billion video collection--now there's something in between. "The point is to bring channels first and foremost into the site," Lovinsky says, "and to create an experience where I as a user can tell YouTube, Hey, these are the things that I care about, this is where my affinity lies, this is the thing I like to be connected to, so I can turn on YouTube and see what's on."

It's a minor change that has major repercussions, not just for you and Your Tube but quite possibly for the entire broadcast industry. It helps solve YouTube's chaos problem, by putting users to work organizing all those videos themselves. But more to the point, there's something else that has channels--oh, that's right, TV! This isn't just a design tweak; this is YouTube attempting to take its place as the next and possibly final stage in what it sees as the lengthy evolution of broadcast media everywhere.

Shishir Mehrotra, YouTube's vice president of product management (after a while, you begin to suspect that everybody at YouTube has the same job with slightly different names), has a good set piece on this subject. It goes like this: in 1988 the No. 1 TV show in the country was The Cosby Show, with an average weekly Nielsen rating of 27.1. (Mehrotra cites the numbers from memory; he gets some of them wrong, it turns out, but his point stands.) Ten years later, in 1998, Seinfeld was No. 1--but it drew only a 21.7 rating. In 2008 the No. 1 show was American Idol, but it averaged just a 15.4 rating. Obviously, No. 1 isn't what it used to be. The audience for TV has become increasingly fragmented.

But cable TV can fragment only so far. You can have a cable channel that's dedicated solely to sports, but you can't have one that's dedicated just to, say, sailing, because the economics don't work. You can't run a cable channel if only 30 people watch it.

But you can run a YouTube channel. YouTube can fragment infinitely, and Mehrotra thinks it will. YouTube is planning on playing the long game. "If our journey is a baseball game, we're not even in the first inning," he says. "We're, like, in warm-ups. You can't even watch YouTube on your television yet. All the channel owners are still producing for antiquated, gate-kept ecosystems. All that's going to change." If YouTube can't beat TV, it's going to quietly, subtly join it--and then it's going to beat the living crap out of it. "About 75% of our time is spent watching brands that didn't exist in 1980," says Salar Kamangar, YouTube's CEO. "We think of ourselves as the platform for the next generation of channels."

With that in mind, YouTube has started getting into the business of producing its own content, just like a TV network. To do that, it poached a man named Robert Kyncl from Netflix and made him its global head of content. Kyncl--whose thick accent (he's Czech), large teeth and furious energy inevitably remind one of Arnold Schwarzenegger--has been forging partnerships with established old-media creators all over Hollywood, offering them cash, low production costs and no hassles. "All content creators, especially the more successful they become in television, the less happy they are with the way their art is treated," Kyncl says. "They're getting notes and creative direction from those who find the audience for them, which is the TV networks. They view YouTube as a place where they can find creative freedom." So far Kyncl has signed deals with Jay-Z, Madonna, Disney, the Onion, Amy Poehler, Tony Hawk and Anthony Zuiker (he created CSI), among others, to build channels on YouTube, some of which have already launched. If it all sounds a bit quixotic--a website for novelty videos, owned by a company without a single content-related base pair anywhere in its corporate DNA, trying to take on all of TV--consider this: right now, the most popular channel on YouTube has 5.3 million subscribers. That's impressive in itself, but even more so when you consider that cable TV's top network, USA, averages only 1.3 million viewers over the course of each day. Granted, USA's viewers spend a lot more time watching USA than YouTubers do on YouTube, but still. The scale is still there.

Though just because YouTube has channels now doesn't mean that its rules have fundamentally changed. The nature of the beast is still the same, and one wonders if the likes of Madonna and Jay-Z fully understand that. For instance: that YouTube channel, the one with the 5.3 million subscribers, doesn't belong to a celebrity or a major broadcast network. It's called RayWilliamJohnson, and it belongs, not surprisingly, to one Ray William Johnson, a law student turned video blogger who does rapid-fire commentary on other people's viral videos.

That's one of the funny wrinkles in YouTube's channel strategy: anybody can run one easily and for free. That puts individual YouTube users on the same footing with celebrities and major networks. They don't even have to produce their own content. Cable TV is very much about buying and/or creating expensive and/or original programming, but more content is the last thing YouTube needs. It's drowning in content, at 60 hours a minute! So a channel owner can create value simply by curating and organizing content, which is something that an individual or a small group can do as well as a network, and sometimes better. Johnson is a good example. Another is the Young Turks, a news channel that reposts and comments on other people's footage and interviews. CNN's YouTube channel has 50,000 subscribers, but the Young Turks, with a fraction of CNN's resources, has 315,000. In the land of abundance, it's the curator, not the creator, who is king, and authenticity and relatability are worth as much as or more than slick production values.

YouTube can't just step in and replace TV, because it's a fundamentally different medium. If it does, the world is going to look very different. YouTube is an inverted, looking-glass version of the media landscape: brands that are dominant everywhere else play like amateurs on it, and amateurs play like multinational conglomerates. The Khan Academy is an education channel on YouTube founded by a former hedge-fund manager named Salman Khan, who makes almost all its videos himself. It has 255,000 subscribers; that's more than the channels of the entire Ivy League combined. The Cartoon Network's YouTube channel has 53,000 subscribers; compare that with Fred, a channel run by an 18-year-old from Nebraska named Lucas Cruikshank, which has 2.4 million. And so on.

That's not to say that a big institution couldn't do what Khan did. It just wouldn't, as Mehrotra is the first to admit: "Imagine [Khan] walking into a studio and saying, 'I want to start a show about math. All I want to do is solve problems. I'm never going to show my face. I'm going to solve every problem in your standard algebra book, then I'll do it in chemistry and so on.' He'd get laughed out of the room." Brands and celebrities occasionally make it out of YouTube--Justin Bieber being the obvious example--but it's much rarer for them to successfully cross over the other way.

This isn't lost on Kyncl, by the way, who likes to matchmake his professional partners with YouTube natives who can show them the ropes. But he has his work cut out for him. Run your eye down the list of the most-subscribed channels on YouTube and you have to go down 14 places before you get to a name that a person on the street would recognize (that name is Rihanna). Something about the air on Planet YouTube is toxic to the professional, corporate, branded way of doing things. It's not entirely clear whether professionals can breathe the air in what has, until now, been an amateur's paradise.

The danger for YouTube is that by trying to beat TV, it will become TV, and in so doing it will lose its weird, fluky, anarchic heart, which is what makes it different from, and in some ways better than, TV. But more likely, YouTube's new strategy will end up unleashing the billion-eyed beast, rather than taming it. It will undoubtedly grow YouTube's audience, but there's no reason to think it will make our tastes less weird and random. If the Young Turks are the future of broadcast news, advertisers may just have to make their peace with that.

Take another look at Nyan Cat. It may just be a cartoon cat in space (with a Pop-Tart body and trailing rainbow), but it has a serious lesson to teach us. If you search for Nyan Cat on YouTube, you'll find hundreds of versions of it, reflecting thousands of hours of work from people all over the world. There's an Indian Nyan Cat (its body is made up of garlic naan instead of a Pop-Tart). There are live-action versions featuring both cats and humans dressed as cats. There are cover versions scored for solo piano, solo violin, a smooth-jazz combo and a full orchestra. There's a Lego version, a 100-hour-long version and a Nazi version (Nein Cat!). There's a Schrdinger's Nyan Cat. There's a Star Wars version in which Luke Skywalker and Obi-Wan Kenobi watch the holo-message from R2-D2, only instead of seeing Princess Leia, they see Nyan Cat. There's a video of a cat watching Nyan Cat, and a video of a cat watching that video and a video of a cat watching that video.

If the studios would have laughed at Sal Khan, what would they say to Nyan Cat? What Nyan Cat tells us is that when you put amateurs in charge of broadcast media, odd things happen, and that's what YouTube does. When all we had was bland, corporate network television, we assumed that we were bland and corporate too. But if we're the ones running the studio, the Nyan Cat will be out of the bag. We're only beginning to find out how weird we really are. Stay tuned, because YouTube is going to show us.

FOR MORE COVERAGE OF THE PHENOMENON, GO TO TIME.COM/YOUTUBE