It’s nearly 11 pm on a Monday night in late July, and we are in the back room of an Italian restaurant not far from the Fort Lauderdale beach. It’s a place he often takes visitors who make the trek from Los Angeles or San Francisco to Mickey Mouse’s Florida homeland for a demo. Oscar-winning visual effects wizard John Gaeta, known for his work on the Matrix and later at Lucasfilm’s ILMxLAB, sits to my right, having joined Magic Leap last year. Former Samsung executive Omar Khan, who is on day 11 in his new role as chief product officer, sits to my left. Everyone is in a good mood because finally, I mean finally, after two years of boastful promises followed by two years of near silence, the company is on the cusp of revealing a headset that actual developers—and any old person in the wild—will be able to buy and bring home.
But it’s unclear, now, whether enough people will be willing to try it. Such a thought would have been absurd just three years ago when Magic Leap was the hottest company in augmented reality, and any interaction with its secretive technology became a status symbol among techies. Yet Magic Leap has promised so much for so long, with no results to speak of, that many of those who occupied that first wave of hype have written off all hope that its infamous, mythical, mixed reality product is real at all, let alone the transformative technology it set out to be.
Abovitz gets it. In the fall of 2014, when Magic Leap’s entire staff could still fit inside his conference room and demos were run on a refrigerator-size metal block nicknamed The Beast, Google led a $542 million series B investment. It was an absurd-sounding amount of money for such an early round of funding—and Google’s CEO Sundar Pichai joined Magic Leap’s board. Suddenly everyone was curious about the mysterious Google-backed tech company with a quirky founder who eschewed Silicon Valley’s norms—a man who planned to headquarter his company in Florida rather than moving to the West Coast.
Where virtual reality surrounds a user in an artificial world, augmented reality superimposes virtual objects into your real-world surroundings. In its simplest form, that means an overlay of information, like a dangling Pikachu in Pokémon Go. But Magic Leap, has sought to give those superimposed objects shape and heft—if you’re seeing something, you can touch it, or move it, or interact with it, as if it were real. Abovitz bragged that Magic Leap had a new super-slick method for superimposing these digital assets —a technique called digital lightfield technology—that was better than anything Microsoft or Facebook or anyone else had developed.
But Abovitz’s grand descriptions always seemed to fall short of explaining how this technology worked. Before Magic Leap had a headset or software or programs, it hired marketers to sell the Dream of Magic Leap, all the while promising that a product was just around the corner. Abovitz dropped mysterious hints on Twitter, hid Easter Eggs inside of old TED talks, and accepted an invitation to speak at the 2015 TED conference, bailing just days before his scheduled talk.
For the past four years, the headsets Abovitz has promised have failed to materialize, and tech prognosticators have begun to question whether people will even want to wear headsets at all, now that they can simply pull up apps on their iPhones to augment reality. The company has guarded its secrets, revealing very little about how digital lightfield technology works, or what its future product might look like. Developers, analysts, and general tech enthusiasts had grown increasingly skeptical that Magic Leap was developing anything worth following at all. Headlines ask “Why do people keep giving Magic Leap Money?” and cry“Believe it or not, Magic Leap says its headset will ship ‘this summer.’” As Jono MacDougall, developer and author of the blog GPU of The Brain, blogged, “It makes us feel like they are in a secret club and they won’t invite us in. It makes us feel like they think they are better than us.” He added, “It makes us want them to fail.”
Yet for all the things that have gone wrong, a few important things have worked out. Magic Leap has now raised more than $2.3 billion, enough money to fuel a lengthy research and development phase. It has built a stable of advisers and investors that include Alibaba executive vice chairman Joe Tsai, Hollywood director Steven Spielberg, and Sir Richard Taylor, who shepherds Magic Leap’s alliance with his mixed reality and game studio, Weta Gameshop, a division of Weta Workshop. It has built an eclectic workforce of around 1,500 employees. Over the past year, under the leadership of new chief marketing officer Brenda Freeman, the company has endeavored to stop playing up the mystery and start telling people about the product it’s trying to build. “Secretive is this word that can be really loaded,” Freeman says. “The idea is to make sure people trust us and we have credibility.”
The first test of this trust is about to occur. Today, the company begins to sell a $2,295 headset called the Magic Leap One Creator Edition in six US cities. Abovitz is certain, or at least he’s really hoping, that once developers begin to play with the company’s invention, they’ll drop their complaints and change their minds about Magic Leap. And then, as they code new games and other experiences, everyone else will too. But Magic Leap’s next phase rests on its ability to do something wholly unconceivable to early Magic Leapers: It needs to succeed as an ordinary company.
THE TL;DR VERSION of Magic Leap’s history goes like this: In 2011, after Abovitz sold his surgical robotics company, Mako Surgical, he teamed up with two others to start building Magic Leap. For the first few years, the “wandering through the desert years” as Abovitz calls them, the company kept a low profile. But the year Magic Leap raised its Google money, it also began trying to seed the promise of what it was building. Abovitz hired a flashy marketer, Brian Wallace, who’d been at Blackberry and helped launch the “Next Big Thing” campaign at Samsung. He began hosting celebrities interested in seeing how his goggles could change Hollywood. A pilgrimage to South Florida became a status symbol among tech and media execs. (Beyoncé reportedly found the demo boring.)
Building a new headset turned out to be much harder than Abovitz anticipated. “I came out of building robots for surgery, so I thought this would be easier than that,” he says now. “It’s easier in the sense that we’re not cutting into people, but at a systems engineering level, this is harder.”
Abovitz had promised not just new hardware, but also new software that could function inside of an entirely new paradigm. What does a desktop look like outside of the two-dimensional bounds of a computer? How do people interact with controls when they’re pointing a ray in the air, rather than moving a mouse around a screen? And, for this audacious invention to take off, Abovitz needed to cultivate a group of developers excited enough about the goggles and intrigued sufficiently by the software that they’d build their own programs for it. It’s hard to win over developers until you have a product and customers to test the tools they create. You must convince them to join you on faith. So no, it wasn’t going to be ready in 2014. Or 2015. Or 2016. Or, actually, 2017.
Meanwhile, Abovitz clashed with Wallace, who had a background leading large national branding campaigns with more predictable product development and launch cycles. Wallace grew frustrated that he felt he was being asked to sell something that didn’t exist. “As the months and years wore on it became clear to me that what he was directing us to say publicly was not going to converge with the realities of the product when it launched,” Wallace now says.
Abovitz says the marketing team wasn’t in line with the company’s culture. In retrospect, he likens it to a type of organ rejection, describing the ethos of the time as tribal. “It was like, which culture is going to win? This splashy big company kind of thing? Everyone else was just like, that doesn’t feel right.” To cut to the chase, he explains, “we were not connecting,” he says.
The last time I visited Magic Leap was the spring of 2016. There was a palpable tension when the marketing side of the business promised I’d see a product by the end of the year. Even then, it felt ingenuine. In my demo, the goggles were still attached to a large computer, and the software crashed after 10 or 15 minutes.
Later that year, Wallace’s contract was “terminated without cause.” Abovitz brought on media executive Brenda Freeman, a former chemical engineer who came from a media background and had been the chief marketing officer of National Geographic Channel. She arrived at a company that “was more focused on the marketing message” than paying attention to what the company was building, she says now. “It was empty carbohydrates,” she says.
Freeman got rid of the large agencies. For awhile, things got very quiet.
TODAY’S MAGIC LEAP offices look a lot like any tech company headquarters you’d visit in Silicon Valley, minus the gourmet chefs. (Breakfast is catered on Mondays and lunch on Wednesdays and Fridays. This qualifies as a perk in Florida.) Walk past two security guards into an atrium where a wall of preserved plants rises two stories. Pass over your ID, and the receptionist will send you up the staircase, past a second security guard to a reception area, where you will be met by an escort.
The differences between Magic Leap and other tech offices are subtle. You’d have to look past the rubber unicorn head and the red stuffed version of the corporate logo, a “Leaper,” hand-knit by Abovitz’s wife, to notice the talking stick, which he had his designers make to help ensure that the white guys and loud mouths don’t do all the talking at meetings. And you’d have to be wearing a headset, as many of the employees wandering through the office are, to see a deer grazing in front of the couch in Abovitz’s glass-walled office. Or to notice the T. rex in the hallway just outside. As I stand in front of Abovitz’s desk, watching the dinosaur stretch his neck, a man walks behind the cartoon character and he is completely obscured.
Over the past two years, Abovitz has attempted to move the company out of the research and development phase, and focus on finally shipping a product. In early 2017, Magic Leap put out a call for developers—Abovitz did this via tweet—and began to seed the company’s developer tools among more than 100 of them, spread out around the world. Then, in December, the company finally revealed its headset with a blog post and a feature in Rolling Stone. This was followed in March by a preview of its software development kit and a website dubbed the “creator portal” with resources to help interested developers.
In the future, Abovitz hopes that developers programming for Magic Leap will be able to have their creations rendered on multiple operating systems. That T. rex outside his office? Theoretically, he hopes you should be able to see it via any computing device you choose, including your phone. It’s a democratic approach to building software that’s a smart move for a company looking to lure developers with the assurance that their work will be seen. In order for the product to become a mainstay, Magic Leap will need to coax them to spend their time programming for its headset with the promise that there will be future customers.
To entice them, Magic Leap is trying to make development dead simple. As the chief content officer, Rio Caraeff, explains: “We don’t have, you know, 50 million developers, and we have to do everything possible to make it easy to build for Magic Leap.” Sure, there will be ambitious creators right from the start, some of whom Magic Leap has partnered with just to help a larger audience reimagine what the technology can do. The Icelandic band Sigur Rós worked with Magic Leap to build an electrifyingly beautiful visual sound experience. But Magic Leap has also provided a way for developers to input a tiny snippet of code into their existing projects and refer to 3-D models to render web pages in 3-D in Magic Leap’s Helio browser. So, you can open a demo of The New York Times in Magic Leap’s Helio browser, just as you might on your desktop. But by adding a small snippet of code that renders a 3-D model, The New York Times can also show you a news photo rendered in 3-D so you can more closely explore it.
In seeding developers, Magic Leap is attempting to steer the design direction of its technology. Sure, 40 percent of the developers who received the goggles early are focused on gaming and entertainment, use cases that have been the company’s mainstay. But Magic Leap has also developed tools for corporate communication (Imagine Zoom, but if your entire conference party were avatars sitting around a digital conference table.) Roughly 10 percent of the company’s existing developers come from healthcare and medical imaging, which isn’t surprising given Abovitz’s background.
I TRIED OUT the Magic Leap One in a 1,000-square-foot faux-living room that had been tricked out in West Elm furniture, and it wasn’t great at first. The headset was beautiful, and unlike others I’ve tried, it felt light on my head. A disc-size battery and computing pack, built like a small CD, fit easily in my front pocket. A main menu popped up in front of me, the field of view large enough that it didn’t seem narrow. But as great as this was, there were glitches. When I tried to use the hand controller to navigate to a football demo, the controller didn’t respond; the experience appeared frozen.
Sam Miller, the baby-faced former NASA engineer who says he stayed up all night the first time he met Abovitz to brainstorm how Magic Leap might work, stood along the side wall, watching, trying to locate the problem. The woman who was giving me the software demo removed the headset, took a look inside, and returned it to me. Miller suggested I change the nose bridge. Finally we got the thing working.
The problem, we figured out, was the fit. Each set of goggles comes with five nose pieces and the ability to add prescription lenses. Magic Leap’s experience is so closely linked to a person’s physiology, these goggles will need to fit perfectly to work. This is a challenge for a company trying to introduce a new type of tech. So, Magic Leap has contracted with former Apple executive Ron Johnson’s startup, Enjoy, which sends customer service people to deliver new tech gadgets and help users set them up. Enjoy representatives will deliver every Magic Leap headset, fit it, and provide a tutorial on how it works.
Once the headset was working, the experiences were creative and compelling. The images were crisp and solid (as solid as virtual reality can be, anyway). With a click of the controller, I pinned a Helio browser on the wall to my left. I opened a Wayfair demo in a second browser directly in front of me. A plush chair appealed to me from the Wayfair website, so I used the controller to drag it directly into the living room and see what it might look like. I finally saw the NFL demo, using my controller to plant the playing field on the ground so that I could better see an interesting move in a game I was watching. And I got to play Dr. Grordbort’s Invaders, a first-person shooter game being developed by Weta Gameshop.
With help and support from a friendly robot called Gimbal, I shot at the menacing robots that were walking ominously toward me until I’d killed them all and a portal opened in the wall. As I was spraying my ammunition, I accidentally hit Gimbal, and I was so caught up in the game and the character that without thinking, I apologized to him.
These experiences are certainly on par with other augmented reality and virtual reality demos I have seen. Are they really mind-blowingly better than the competition? Not yet. But Magic Leap does have a product, and despite its naysayers, it’s very close to being in all of our homes.
WHEN WE FIRST met in 2016, Abovitz told me he envisioned a great unveil for Magic Leap. He said he hoped it’d be like Charlie and the Chocolate Factory, that a few lucky people would find golden tickets and show up for an event like none before. It hasn’t happened that way. Instead, Magic Leap’s great unveil has happened a lot like the product releases at other large tech companies: a spattering of press, centered around an embargoed rollout.
This has left many reviewers disappointed. They complain that the device’s field of view is not that much larger than Microsoft’s HoloLens: In other words, Magic Leap’s fancy goggles are just another version of the status quo. A video of a Magic Leap experience that began to circulate in July showed little jittery rock-throwing creatures, far less magnificent than the whale rising out of the gym floor that headlined one of the company’s original concept videos. The Apple blogger John Gruber called the promo disappointing, saying, “I’ve long been suspicious that the reason Magic Leap is so secretive about their actual technology is that it’s nowhere close to what they promised in their concept videos. This seems to confirm it.”
Abovitz calls this his “tv-on-the-radio” problem. He took to Twitter shortly after the video’s release to respond to critics. “Most people just got it once they first saw TV. Same with @magicleap,” he tweeted. Abovitz believes that the only thing that will assuage the public’s doubts about the viability of his product is for those haters to see it, touch it, engage with it, and come away believers—that no amount of promo and marketing will convey what Magic Leap is attempting.
I will agree with all the people who have seen Magic Leap and report that the technology paired with the choice of content really are different and better than other things I’ve seen. The company says it has already given 10,000 demos, and it will start a roadshow this fall to bring Magic Leap systems to college campuses and other places where the general public can try it.
But good technology doesn’t guarantee that a company will redefine a field. Many tech companies manufactured early versions of PDAs before Apple produced the iPhone. Furthermore, no one is sure if people other than gamers and perhaps folks in corporations will want to wear headsets when they could just hold their phone screens out in front of them for an augmented view of the world. While Apple’s two-dimensional iPhone approach to augmented reality is hardly as mind-blowing as a hands-free headset, it’s immediately available to everyone and thus practical. The company can leverage its 20 million developers and all of its users to build a range of AR experiences that could later transfer to any device we may adopt.
Magic Leap may not have 20 million developers yet, but it has vast amounts of cash. It has dozens of patents. It has patient investors. And it has a founder who remains resolute in his ambition. “We didn’t just shoot off and do Apollo and land on the moon,” he tells me as I’m getting ready to leave. “It was all these incremental steps, and if we hadn’t done it that way, there would have been no learning.”
Maybe he’s right—that all the hype and marketing belies the fact that Magic Leap is just a sort of normal tech company, applying money and talent in equal measure in pursuit of what is, they hope, a big enough idea. It’s no chocolate factory. There are no Oompa Loompas. There are just engineers and designers, weary from rushing to launch their first product, hoping developers will fall hard for Magic Leap One Creator Edition. But if they don’t, the company will do what healthy, well-funded companies do: try to improve it. Already, in another part of the building, a team of engineers is developing Magic Leap Two. Release date: TBD.