Magic Leap’s first-ever L.E.A.P. developer conference in Los Angeles this week was a chance for the company to finally stop the naysayers. Decried as over-funded vaporware for years, the event represented a unique opportunity for Magic Leap to show that it was the real deal, and that it mattered for the future of visual computing.
The results of those efforts were decidedly mixed. The company was able to show off its Magic Leap One headset to a much wider audience, complete with demos built by a number of high-profile partners, including Lucasfilm’s ILMxLAB, Insomniac Games, and Weta Workshop. But attempts to lay out a coherent vision for Magic Leap’s future were muddled by executives waxing about magic, spirit journeys and goats.
Starting From Scratch Isn’t Easy
When big tech companies like Google, Apple or Facebook hold their annual developer conferences, they tend to use those events to speak to two distinct audiences: Developers looking to build apps for their devices and platforms, and consumers at large looking for some clues for what’s next in tech. That’s why developer conference keynotes often include some high-level talk from CEOs about their companies’ plans, followed by a bunch of product news, and finally some updates on developer tools.
Magic Leap had some good reasons not to stick to this script. The company just started to ship its Magic Leap One headset a few weeks ago. With a price tag of $2,295, the Magic Leap One is decidedly not a consumer device, but squarely targeted at developers. And with so much of Magic Leap’s work shrouded in secrecy for so long, the company was likely better suited to clear the air, lay out its roadmap and vision and ultimately explain itself to the world — a coming-out of kind, if you will.
However, Magic Leap decided to go a different route. When its CEO Rony Abovitz took the stage at the beginning of the conference’s opening keynote, he vaguely framed the company’s technology as a remedy to political and other divisions plaguing today’s world, postulating: “Our new medium of spatial computing is fresh. It doesn’t carry the baggage and negative headlines that are dominating the news today.”
Magic Leap chief marketing officer Brenda Freeman picked up on this idea, declaring: “We want our platform to be a sanctuary,” and emphasizing the company’s focus on diversity and inclusion, with a goal of eventually having gender parity among attendees of its developer conferences. She didn’t say how the company was going to achieve this goal, and a Magic Leap spokesperson declined to comment on the actual percentage of female attendees during this week’s event.
This wasn’t the only commitment to admirable goals without actual details: Magic Leap’s Chief Content Officer Rio Caraeff later committed to investing in content from indie developers as well as women and minorities, but didn’t put any dollar figures to those promises. A spokesperson declined to add any further details.
Fighting fascism in the Magicverse
Instead, attendees got to hear a lot about the role Magic Leap aims to play in some distant future. “The conversation that we are starting here goes on for decades,” proclaimed Abovitz. He went on to talk about Magic Leap’s idea to build city-wide AR information layers — an interesting idea, but not anything we should expect to materialize any time soon, as he freely admitted, while inexplicably bringing up the spectre of fascism:
“There is gonna be other companies proposing alternate x-verses. There will be all these competing universes and systems, like there is competing governments. There is monarchy, there is democracy, there is fascism, and there is progressive, liberal thought. And these holistic systems will be how governance might happen. How many people live their life. And this is not tomorrow, this might be a few years from now. But you set the stage for that right now.”
Magic Leap CEO Rony Abovitz discussing what his company calls “The Magicverse.”
Abovitz seemed to sense that there was still some work to be done before Magic Leap could defeat fascism in the Magicverse — and he promptly proposed to host a Burning Man-style gathering to get it done:
“In order to continue this conversation, we probably kind of need a Burning Man in the desert. Like a kind of Magicverse event where we hang out for a few days, do a vision quest, build weird things. So more on that later, can’t commit to it 100%, but we think it’d be cool.”
There were some other odd moments that morning. A third-party developer read a poem to the audience. Science fiction author and Magic Leap chief futurist Neal Stephenson talked close to 10 minutes about virtual goats. The three-hour keynote didn’t feature a single live demo.
And when it was time to talk about the Magic Leap One’s Lumin OS, a slide showed off icons for apps from Netflix, Google and Nintendo. A spokesperson later confirmed that these were added for illustration purposes only, and don’t represent any actual partnerships — a pretty big snafu for such a high-level event.
The potential and limits of Magic Leap’s hardware
That’s not to say that Magic Leap doesn’t have any impressive partners. In addition to the aforementioned ILMxLAB, Insomniac Games and Weta Workshop, the company also used the event to highlight partnerships with AT&T, Wayfair, Wingnut AR, Epic Games, Unity Technologies, and others.
A rendering of WingnutAR’s Magic Leap demo.
And while the keynote was devoid of actual demos, attendees had the chance to try out a number of these games and experiences first-hand throughout the conference. Many of these demos were abbreviated versions of actual games, but even just a few minutes of some choice titles showed some of Magic Leap’s potential, as well as some of the shortcomings of its current-generation headset.
Anyone who has ever tried Microsoft’s Hololens, or any other first-generation AR headset, immediately recognized that the Magic Leap One is a lot more advanced. Digital objects overlaid over the real world aren’t as translucent as they used to be on previous-generation headsets, and instead actually cover real-world objects behind them. Magic Leap also seems to have gotten pretty good at surface detection, allowing characters to jump on and off tables, couches and more.
A promotional rendering of Weta Workshop’s “Dr. Grordbort’s Invaders.”
The most obvious technical limitation is the headset’s field-of-view. The Magic Leap One can only use a small window right in front of your eyes to display digital objects and characters. That’s most apparent when you find yourself across from a life-sized AR character, and in turn forces many developers to think smaller.
The Magic Leap One also has some issues with objects in close proximity. Lift a digital object up close to your face, and it disappears in thin air. A spokesperson said that the hardware was set to a near clipping plane of 14.57 inches to minimize discomfort.
The biggest issue is that for all of its technology, the Magic Leap One is still fairly dumb. It can recognize surfaces across walls, ceilings and furniture, but it doesn’t actually know what any of the objects in your home are.
The result is that games and apps developed for Magic Leap can’t actually interact with the real world, only embellish it. Multiple games shown at the conference opened up portals to virtual worlds in walls and floors, complete with characters spilling out of those words into your living room. That’s fun at first, but quickly feels like an AR parlor trick.
ILMxLAB’s “Star Wars: Project Porg” showed where things could go by incorporating a number of objects, including a Sonos speaker and a TV screen, into the experience. However, this level of interactivity was custom-made for a carefully configured demo space, and won’t be included when “Star Wars: Project Porg” ships on the headset later this year.
A roadmap that doesn’t go very far
Magic Leap is working on object recognition technology, according to a software roadmap shared during the L.E.A.P. opening keynote. The company didn’t share when that technology will ship, how advanced it is, or what kind of applications it will enable.
Likewise, Magic Leap also didn’t give any indication for its future hardware plans. Right now, developers are left guessing when there will be a consumer version of Magic Leap’s headset, how much it will cost, and what it will be capable of. Instead, Magic Leap only chose to share a few select updates for the very near future, which include software features it plans to launch during the remainder of the year.
This lack of transparency is a problem for the company. Magic Leap could get away with being the crazy thing that everyone doubted while it was still operating in stealth mode. Now, it needs to show that it is more than just hype, and actually has a technology roadmap that can turn it into a dominant player in a space that is already attracting competition from industry giants like Facebook, Apple and Google.
To do so, Magic Leap needs to stop behaving like a bunch of guys on a vision quest, and start acting like a company that can actually deliver. In other words: It’s time for Magic Leap to get real.