In late 2013, Clay Bavor began experimenting with teleportation. He paired an Oculus Rift headset to a robotic arm, upon which he mounted a couple of GoPro cameras. When he moved his head, the thinking went, the cameras would mimic the movement, acting as a second pair of eyes. If it worked, he'd be able to "teleport" himself (or his eyes, at least) a few feet away. He still has a video of the first time he ever got it running: There's Bavor, tall and thin in a t-shirt and jeans, standing among the contraptions with the Rift on his face. He reaches out his arm, waving his hand in front of the cameras at his side while simultaneously seeing it in front of his face. "Whoaaa," he says to himself. "This is crazy. This is like the craziest thing I've ever experienced."
The so-called Teleportation Robot was just one of Bavor's many side hustles at Google. Technically, his job was to lead the company's apps teams---the folks who make Gmail and Drive and Docs. But he'd been enthralled by VR ever since he first tried the Rift, and so for about a month, he spent his 20-percent time (that window Google famously gives employees to explore outlandish new ideas) playing with ways to enable sustained eye contact on a video chat. He started with reflecting mirrors, 4K TVs, and teleprompter rigs, but his ambitions quickly leaped another several levels. Forget eye contact, he thought. It should feel like you're there. He built the Teleportation Robot, and started subjecting colleagues to robot-enhanced meetings. He'd bring the Rift and a Mac with him everywhere, giving demos to anyone and everyone. Bavor developed a reputation as Google's VR guy long before Google had a job opening for "VR guy." Now virtual reality is his full-time job.
The boyishly handsome guy with a fondness for gray hoodies over gray t-shirts is Google's VP of VR, the leader of a team that numbers "many hundreds." They've taken over a building on Google's campus, not two years after Sundar Pichai casually mentioned its first VR product on stage. That product, Cardboard, a dirt-cheap makeshift headset, can't compete with the (now finished) Oculus Rift or HTC Vive in picture quality, and the buzz around it pales in comparison to its competition. But that doesn't bother Bavor. What he's after instead, what Cardboard has instead, is distribution: Google's given away and sold thousands of headsets, brands hand it out to encourage people to try their own VR apps, and everyone from The New York Times to Coachella is distributing Cardboard to their customers. "It's let us engage in VR," Bavor says, "and learn, and bring a lot more people into the story of the beginning of VR than would otherwise be there." The five-millionth Cardboard shipped out in January 2016, when there still was no Vive or Rift or Playstation VR. Google's virtual reality is for everyone, and it's already here.
Make no mistake, however: Google’s interest in VR goes far beyond Cardboard. Google is an investor in Magic Leap, the super-secret augmented reality platform. Google has also acquired Tilt Brush, the paint-in-free-space app that tends to be the one that really gets people’s brains churning with the possibilities of VR. But where other VR companies are aiming for perfection now and ubiquity later, Google is turning that formula on its head: get VR into as many hands (and onto as many faces) as possible, as immediately as possible. Then make it great. Inside Bavor’s own team, they have big plans for getting there. Plans that touch nearly everything Google does, and a hundred other things besides. Plans that include your smartphone, and plans that go far beyond.
A few months after Bavor started showing off his robot, David Coz, an engineer in Google's Paris office, came to Mountain View with a much simpler teleportation idea. The first-ever prototype of Google Cardboard was only slightly clunkier than the final version: Coz had a hacked-together holder for your phone along with some lenses and nifty software that gave you stereoscopic vision. He started showing it around, eventually getting it on Bavor's face. "I was just like, this is amazing," Bavor remembers. "I love this, we need to make it." Pichai and CEO Larry Page agreed. They decided to launch it at the upcoming Google I/O developer conference. That was the good news. The bad news: I/O was eight and a half weeks away.
About a dozen Googlers ditched their real jobs, their families, and their circadian rhythms for the next two months and built Cardboard into a launchable product. Pichai's announcement was casual, almost mysterious in its nonchalance, but buzz built fast in the hours after I/O attendees filed out of the Moscone Center hall and started putting together the flat piece of cardboard they'd been handed. That same day, other companies started asking for Google's help in making and selling their own Cardboard designs. Google also gave every attendee a smartwatch that year, but the only piece of hardware to have lasting impact was Cardboard. Virtual reality was the hit of the show.
From the beginning, Bavor was a natural fit to lead the effort. For one thing, it was clear from the get-go that VR would ultimately be a Google-wide effort---even the very first app, at I/O, included work from YouTube, Maps, Earth, Spotlight Stories, and more. Bavor has an aptitude for running teams that integrated across Google. "He's just known for doing that really well," says Andrew Nartker, a longtime collaborator of Bavor's and now a product manager on his team. Everyone also knew about the experiments, the Teleportation Robot and many others in the ersatz lab Bavor had set up. Then there was the fact that, well, Bavor's kind of made for this job. You could actually argue his whole life's been building up to this.
As a kid, Bavor didn't exactly spend his days imagining his career in virtual reality. Growing up in Mountain View, CA, maybe five miles from his current office, he was mostly into art. He'd thumbed through the pages of a book his dad owned, Photorealism by Louis Meisel, and fell in love with the incredibly realistic paintings of artists like Chuck Close."You see these paintings, and you're like, oh my god, that's an amazing photo," Bavor says. "And then you're like, oh, that's not a photo." A gifted drawer, he was totally uninterested in abstracts or impressionistic masterpieces. All he wanted was to make you think his drawing was a photograph. "My mind was just blown by the idea that you could, using things that weren't real, make things look real." The closest thing to virtual reality at the time was Nintendo's ill-fated Virtual Boy, but Bavor was already training for his future.
His family had an Apple LC II at home, which became Bavor's window into an entirely new kind of faux realism. "The first time I saw a scanned photograph on a computer," he says, "I was like," and here he leans close and whispers excitedly as he always does when he's recounting some mind-bending discovery, "Oh my god. It's something real. On the computer." He became infatuated with rendering and 3D animation, using software like Ray Dream Studio and Strata Studio Pro before winning a $7,500 copy of ElectricImage, a high-end animation software used in movies from Terminator to Star Trek, in a raffle that may or may not have been fixed in his favor. (Long story.) He'd take his dad's guitar, or his Discman, and spend days making sure every button, every shadow, every curve looked just right. Then he'd show his dad, and rejoice when the older Clayton Bavor complimented his son on a lovely photograph of his guitar. By the time he was in eighth grade, he'd worked up the courage to tell his dad, a doctor in a line of doctors, that he didn't want to go to med school.
He studied computer science at Princeton, and during his time there built the web portal for the entire student body, along with an auction system called Tiger Trade. Nobody asked him to; he just did. "I get very upset," he says, "when I see something that should exist that doesn't." In another life, Bavor says he might have left school and become a neuroscientist or a special effects artist. In this one, upon graduating in 2005, he spent six years working in every part of Google's ad system until he figures he understood it better than almost anyone. This is a theme for Bavor: He decides to do something, and he does it completely. Until he's the best at it. Then he moves on.
Even as tech became his professional life, Bavor maintained what he calls his "eccentric interests." They're more like one-man self-improvement workshops. One year, he decided he was going to perfect the making of peanut butter toast. Seriously. He's got the recipe down: "Thick slice of good bread, toasted golden-brown. Layer of butter, applied like you would apply slices of cheddar cheese. Layer of peanut butter, layer of bananas. Drizzle with maple syrup, sprinkle with sea salt." He also makes an incredible grilled cheese sandwich, and a bacon, egg, and cheese on a Tartine Bakery croissant that you'd kill your family for. Now he's working on frying stuff. "If I'm cooking, he says, "I will cook the same thing again and again until it is perfect." This kind of blindered focus is his happiest, most natural setting. When he started working on Cardboard, all the while still running product and design for Google's apps team---his job post-ads---it was pulling him in too many directions.
In January of 2016, after 18 months of running Google's Cardboard project while still in charge of the apps team, Bavor got a new job running Google's also-new VR team. Apps have a new boss, Diane Greene, and VR became Bavor's only project. "Google's known as the cool Cardboard company," says Brian Blau, an industry analyst at Gartner, "and otherwise, people probably think they don't know what they're doing." The shuffle was a signal from above that Google was getting serious about VR, and it gave Bavor a chance to be as single-minded at work as he is in the kitchen. He found himself leading a big, growing team in a nascent industry with few right answers and no winners, but impossibly high stakes. VR is coming, and it's coming fast. Analysts disagree on the numbers, but they're all huge. By 2026, one wrote recently, VR could be bigger than TV. Everyone's interested, everyone's watching, everyone's working on something. It comes down to this: You can't get VR wrong. And you might only get one shot.
It's a rainy March day when Bavor and I first meet, inside his small office. He's in his usual uniform: gray hoodie over a gray t-shirt with a line-drawing of a Cardboard headset, jeans, New Balance shoes. He's just getting over being sick, but perks up immediately as soon as he starts talking about the four Avegant Glyphs he bought for his team to check out. He likes it, especially because it's so easy to use. "The interface is just HDMI!" He reminisces about CES, all the crazy virtual reality experiments he saw, like the one that involved a mattress and did something he can't remember. But he gets most excited when he remembers the first time he saw the Oculus Rift, in the summer of 2013. That was when he knew truly great VR was actually possible. The Rift wasn't great then, he says, "but the big things that need to be there are there."
In the long run, VR doesn't really work when it's impressionistic, or kind of almost close to great. Pretty good VR is only cool until the novelty wears off. Then it's just shitty VR. It sounds odd to say Bavor understands this, given that his team's only output so far has been by far the least technically impressive virtual reality platform on the market. He does understand it, though. It's what drives him and his team forward. But he's also practical: He's read the journals and the magazines, he's tried the demos, and he's firm in his belief that you can't do perfect VR yet, at least not in a way that's accessible to everyone. And being accessible to everyone, for Bavor and for Google, is the whole point.
From the very beginning, Google's goal has been to get its products in front of as many people as possible. Search was about making it easy for anyone to crawl the entire Internet. YouTube was initially conceived as a much-needed place to store videos online. Chrome was the browser for everyone; Chrome OS the computer for everyone. "That philosophy has kind of been a consistent theme of the best things Google has done," says Mike Jazayeri, the product director for Google VR, and they planned their VR path the same way.
The true genius of Cardboard was that you probably already own the most expensive and complicated piece of the equation. You probably have it with you right now. "You don't need to make a decision about where to go use your smartphone," Bavor says. "You don't need to go back into your office or the room you've set up for your smartphone to use your smartphone. And you don't have three cables connecting it to a giant box." All that is why, even as higher-end, more dedicated VR platforms start to come out, your phone is still the centerpiece of Google's short-term plans. They asked themselves at one point, why don't we just build the perfect headset? Get the best displays, processors, everything. It would cost, say, $100,000. It would be incredible. "Sure, we could go do that," Bavor says. "But it doesn't lead directly to bringing this technology---and the best seat in the house, anywhere---to the world."
Everyone at Google talks with pride about their partnership with The New York Times, which brought VR into 1.3 million homes one Sunday morning. "You could say that they invented mobile VR," Blau says, "or that they made it popular in a way that no other company could have done." It matters to Google that it was the company behind virtual reality's first truly mainstream moment. In fact, Jazayeri beams as he notes that 80 percent of Cardboard app installs come from outside the US. While Oculus, Sony, and HTC were building the best platform they possibly could and planning to sell them to deep-pocketed gamers until the price of great tech comes down, Google has resolutely held that it's better to make something everyone can use, and then make that thing better.
Google also exploited a timing advantage: It's still so early in the development of virtual reality that hardly anyone has tried it, and it still doesn't take much to blow people's minds. Every time he gives a demo, Bavor says it goes the same way. The user looks at the piece of cardboard, eyebrow raised, like, this thing can't possibly be any good. They pick it up---lame. They get a little closer---lame. A little closer---lame. It hits their eyes---oh my god. He loves that it goes this way. He even designed Cardboard to elicit that reaction. "The magic in it," he says, "is the gap between what you expect from it and what it delivers." That's why it's named Cardboard, not Paperscope or any of the litany of other names Google considered. Maybe it's not great, but it's better than you thought, and that's all it takes to make you want more.
In the next couple of years, your phone is only going to get better, faster, and higher-res. Bigger, too, probably. It'll start to be built with VR explicitly in mind, incorporating head-tracking software, 3D audio, and maybe even Google's remarkable positional-tracking and world-mapping Project Tango tech. (And apparently much bigger batteries.) Lenovo's already announced a Tango phone, and Bavor and Jazayeri both hint more are coming. Billions of people own smartphones, billions more will soon, and they're getting remarkably more powerful remarkably quickly. Why wouldn't Google ride that wave? Eventually, though, they're going to want to go further. Much further.
As Google's virtual reality team has grown, it's had to grapple with an important question. It's more like a thousand important questions, really, but one in particular: where do you start? VR is so new and so rich with possibility that Bavor likes to respond to people who ask "what's the use case for virtual reality?" by asking about the use case for regular reality. It's that big. And because there are so few right answers, so few hard-and-fast rules about what works and how in virtual reality, every decision feels huge. "Everything that happens in VR is downstream of this moment," he says. "So I remind my team, hey, we're going to do the equivalent of deciding that the way you close a window on a Mac is by clicking a box in the upper left of the window. In VR."
He has this metaphor he uses a lot, about the Dark Cave of Possibilities. Google's in this giant cave, full of paths and roadblocks, rapidly shining a flashlight to illuminate as many things as possible. "It is so new that everything seems either impossible given current technology," Bavor says, "or barely doable." There's a group within the VR team that builds two new apps, every single week, just to see if quilting or gardening or drumming is fun in VR. Bavor's personally curious about haptic feedback, and thinks it's really important to the future of VR, but doesn't think it's possible yet. Same with productivity, text input, all that stuff---until screens get better it's just not really worth even trying. "We'll worry about that in a few years," he says. He's more concerned with figuring out how to do all the things he thinks are already possible. There are plenty to choose from.
My last meeting with Bavor took place in a large blue-walled conference room in Building 40, one of the original Googleplex structures, which is now mostly used for fancy events and executive meetings. Bavor's here because his next meeting is with Pichai, to catch up on VR plans. They do this often. That's because virtual reality is more than just a Google product, he says. "It's everywhere. And it's not just my team. It's the YouTube team. It's the geo team. It's the Maps team. It's the search team. It's the Android team." The vision Bavor sets and pursues is poised to be as important to Google as search---a new technology and interaction method that spans everything at Google.
Two overarching ideas define Bavor's vision for VR. First, his enduring fascination with perfectly capturing and depicting reality. "It was very clear to me that people would be interested in experiencing real-world content," Bavor says. "Teleportation, time travel, being on stage with your favorite artist, seeing a place before you're actually there." This is the thinking that led to Expeditions, an early Cardboard project that let teachers take students on guided field trips to anywhere on Earth. Or, you know, not on Earth. "Underwater is a really popular" field trip, says Jen Holland, a program manager on the team. "Underwater and space. Because for most kids, they're not going to go snorkeling with great white sharks." Bavor wants the next Planet Earth series to be shot in virtual reality, so that he can sit cross-legged in the African savannah and live among a pride of lions. He gets so excited thinking about it that he slips out of his chair and onto the floor of the conference room, holding an imaginary headset to his face, staring open-mouthed at the imaginary lions.
YouTube always seemed like a natural place to host all this content. But how would you capture it? After doing some research into panoramic and stereoscopic cameras, a new camera design just kind of came to Bavor one day, as things tend to. ("He's literally inventing in the middle of [product] reviews," Jazayeri says, "which is incredible.") He drew up a design, did the math, and decided this was possible. He then met with Steve Seitz, a University of Washington Professor, and showed him the blueprints. Seitz told him he was crazy. "And I was like no no no," Bavor remembers, "I think it's really possible. It's hard, but here's how I can prove to myself it's possible." A couple of weeks later, Seitz emailed him. "I've been thinking about your camera," he said. "I think it's barely doable." Not only that, but Seitz and his team wanted to come to Google and work on it. "And not that many weeks later, they'd figured it out. That was Jump, the 16-GoPro rig and some wild processing algorithms that suddenly became the first best VR camera rig on the market. Long-term, Bavor says, 360-degree video of the real world is only a tiny sliver of what's interesting about VR. But it was the easiest, most obvious way to give people a taste of the possibilities.
The second idea that keeps Bavor's mind humming is a longer-term one: How do you go beyond reality, removing its boundaries to give people a truly free place to express themselves? After all, actual reality does a pretty good job already with actual reality. (Bavor says a few times that Mother Nature is really good at rendering.) Only doing real-world stuff in VR "is like saying, hey, we have this amazing special effects system," Bavor says. "Let's... put some more chairs in the room. No! How about a dragon? Or a triple rainbow? Something cool!" He wants to see what the artist Jeff Koons and the sculptor Anish Kapoor could do when they're working not with clay and balloons but light and chrome. That's why he approved acquiring the Tilt Brush team without ever even trying it: "It's version 1.0 for this tool for getting ideas, and objects, and events, and experiences, out of your head and into something that you can see and walk around and bring others into."
Bavor is unsurprisingly coy about his team's specific plans. But there are plenty of believable rumors, like the one that Google's working on a much more advanced, Gear VR-style headset to launch this year. Or the one about the "Android for VR" that might be coming soon. Don't forget about Magic Leap, Tilt Brush, and a host of other high-end experiments taking place every week inside the VR building. This is Google: they don't do small-scale things. All Bavor will say is that Cardboard's nowhere near the last virtual reality project you'll see from Google. Then he leans over, conspiratorially, and whispers again. "I would really like photorealistic virtual reality to exist."
How we'll get there, Bavor doesn't know. He has wild ideas about a spherical light-field camera "that basically records every incoming photon from every direction. And then if you could digitize all that information, and you have on the headset side something that could perfectly re-emit those photons, it would be functionally equivalent to actually being there." Jump is moving in that direction, but it's only just beginning the journey. On the headset side, they're pushing against the limits of resolution, dynamic range, and more.
Perfect VR requires what sounds like a daunting wish list, but Bavor sees it right around the corner. True photorealistic perfection is a ways off, yes. "But for an experience that feels vividly real," he says, "where you're going to be scared of those lions, it's going to happen a lot sooner than people realize." Less than five years. In ten? "I think you'll be approaching...I'm not sure if I'm seeing a real reality or a simulated reality." After that comes the real prize: augmented reality, layering digital content over the real world. Don't forget: before there was Cardboard, there was Google Glass. Google knows better than anyone how powerful that tech is to get right---and how easy the feeling is to get wrong. Maybe they're smart to take this one slowly.
At the end of our time together, Bavor cuts himself off---he has to go meet with Pichai, to catch the boss up on what's happening in the VR team. The group has operated mostly in secret until now, but that may be about to change. It might soon be time to see what Google can do with more expensive materials. "We've been hiding behind Cardboard," Bavor tells me with a mischievous grin. "We have a whole building of people here, and they're not working on thicker versions of Google Cardboard or Cardboard with recycled material." As he walks out, I check the date. It's eight and a half weeks to Google I/O. That's plenty of time.
