I had an interesting conversation with Magic Leap CEO Rony Abovitz on Oct. 9 — the first day of LEAPCon — after filming a 30 minute presentation about their first marquee game, Dr. Grordbort’s Invaders. During the second day of the event Magic Leap hosted a 3-hour keynote which revealed, among other things, a plan to support two controllers with Magic Leap One in 2019.
After the Grordbort presentation on Oct. 9, though, Abovitz agreed to the first on-the-record interview with me. I’ve decided to publish the entire transcript of our conversation so you can decide for yourself what is of interest. I edited out only “ums”, sentence fragments and overly repetitive phrasing.
Hamilton: I’ve obviously been to all the Oculus Connects and it is weird to see a sort of reboot almost for AR, that’s what this conference feels like to me…the ability of devs to get excited for their projects. But it’s a long road still ahead, and I guess I’m wondering what can you say about Magic Leap’s ability to be there for the long haul for devs. I think devs need to know that Magic Leap is going to be there for years to come building new products that are even better than the ones we already have and I’m wondering if you could explain that.
Abovitz: All of the effort we took in building the company in the first place — we set up our own production factory, we raised a certain amount of capital — was to give us the longevity, and sustainability and the basic investors and support to be around for a while so we didn’t build it to do one thing and that was it. We didn’t build it just to do Magic Leap One. Our teams have been working on ML1 and multiple generations afterward and investing in all the technologies that might be showing up two years, five years, even 10 years from now. So part of what we’ve done is invested in the immediate, but also that long road — built up a significant IP portfolio. We’re also working with people like Unity and Unreal so that the skills you build on Magic Leap are extensible so tomorrow we’re going to talk about something we call the “Magicverse” which is like we want things to be XR portable so we don’t like the idea of a completely closed ecosystem.
Hamilton: Like WebVR?
Abovitz: You should be able to build something and be able to interoperate with a phone, a tablet, a VR system, other future devices — so you can imagine you might have a creative world within your building and it exists and there’s different ways to see into that and different ways to consume that. So we want to also make those tools so that you can build a persistent, digital, like geo-spatially locked experience. Our stuff is rooted in the physical world and then it springs into other worlds. As we keep changing devices — and you have different kind of devices — there’s still a way to experience that. That’s what I think of as XR compatibility so the whole idea we want to make people feel like there’s a wider interconnected world, we want to be part of it and we want to link you to other components. We want to build interoperable tools.
Hamilton: Interoperable tools inside your own ecosystem right? We’re not talking about tools to talk to…well you talked about tablets and other things?
Abovitz: Yeah tools to talk to other devices too.
Hamilton: What about VR headsets then, I have to ask.
Abovitz: Sure.
Hamilton: Magic Leap is going to support VR headsets?
Abovitz: So think about — and tomorrow we’re going to talk about this — but our company is built on four north stars. One of them I call sensory field computing which is like light field, sound field, tactile/haptics….another north star we call lifestream which is all the data that you experience and the data of the world around you and how that needs to be protected and how that informs a lot of those things in software which is really unique because it sees what you see, it hears what you hear, it is where you are. Our third north star is human-centered AI because there’s something unique about being able to understand where I am and what I’m looking at, what’s around me and then informing an AI with that very unique and individual thing versus AI which is just fed photos and videos so we’re a unique human-centered AI and those three together combine into something we think of as like a magicverse. We think our set of magicverses are related to generalized xverse. So many different companies create different digital universes or universe types that have different social systems, some of them may be centrally controlled, we want ours to be more decentralized more distributed, more control to developers and users. It is almost like ‘do you live in a monarchy or a democracy?’ … The magicverse which can interoperate, we can have passports in and out of other worlds. We want our sensory field devices to be the best ways to consume them but not the only ones. I think that opens up the world for developers…”
Hamilton: Let me ask another question related to the Facebook one. When Facebook bought Oculus a lot of people reacted such that they realized VR was gonna happen, that spatial computing was the next phase — that purchase kind of spurred a lot of people to believe that that was actually happening.
Abovitz: Because their scale and…
Hamilton: Yeah, the scale and investment is all here. That this is actually going to happen the moment that happened…the problem is a lot of people really lost faith in that startup when that happened. They got scared that Facebook was going to control too much information.
Abovitz: Right, the values of a startup versus the company that bought them.
Hamilton: Yeah, and so I’m guessing what does Magic Leap stand for as a company in comparison to that story and what do you expect to stand for long term?
Abovitz: I don’t want to talk specifically about, unlike others who don’t mind trashing us in the press all the time, I don’t really want to talk about any particular company, but we definitely do have a series of values at magic leap — we call them the magic 7 — and every employee in the company….we believe you have to be socially rooted cultural values and we’ll talk a little bit about that tomorrow at the conference but our goal is I think we are on track for a private company that has nowhere near the funding of the big companies. They’re gonna invest $10-$20 billion…so people go ‘you raised a couple billion.’ Sounds like a lot? The big companies are going to put 10x into what they’re doing. But they’re much more inefficient than what we could do. I’d say our company — if we get the support of developers — we can be a public company. We can be self-sufficient. We don’t have to be acquired by anybody. We don’t have to compromise on our mission and values and north stars. That I would think would be the best thing for the creative community. If we could stay our course, go public at some point, and then you have your own self-sustaining capability. That’s very different than if somebody else takes you over and they kind of impose their philosophy, their structure, their business model. We don’t have a business model that needs to take data from people. Our business model can be built in a much different way, in a much more open distributed way. To me that’s super important.
Hamilton: I guess I’m wondering if you thought the developer community — or the people who are buying your headsets — realize that’s what you represent.
Abovitz: They did not yet, they still don’t. I hope at this conference we can communicate more of that. I think when a lot of folks who leave here they’ll understand more about our vibe, but it is going to take us time to express that, I mean we’re only two months old into the world. We’ve been incubating away and doing our thing and I think you don’t just think about Magic Leap as tech but it is like the culture we’re trying to create with developers, philosophy as a company which is really different I would say from other company philosophies and my job is to keep steering us in that way cuz a lot of times you have all these great values and then they get compromised. So to the extent, how do we scale how are we sustainable and how can we get to a certain size that we are around for decades that we can still have a good reason to sleep at night and people are still excited by us and we haven’t given up everything we believe in to get to that scale. That’s going to be our challenge and my constant thing is to keep gut-checking us about what we’re doing and why we’re doing it, cuz we’re one of the last independent companies in the space of a certain size.
Hamilton: And I think that’s significant and I don’t think people realize that’s what you represent, so that’s something interesting that I wanted to get out there, or at least express, or at least understand better. I’ll just bring up a comment that I heard at the first Oculus Connect.
Abovitz: I have never been to one of those.
Hamilton: A Magic Leaper was at the first Oculus Connect and I asked him, you know, tell me what you’re working on. Couldn’t tell me anything.
Abovitz: Ok
Hamilton: But said the words “fuck pixels.”
Abovitz: Ok
Hamilton: I thought that was a very memorable quote.
Abovitz: That’s kind of a weird [laughs] I don’t know who was there.
Hamilton: It was a very memorable quote and I’m wondering if that’s a spirit inside the company in any way.
Abovitz: I would put it this way. We’re not chasing pixels on the screen. The way I think about resolution is ‘what is the arc-minute on your retina?’ So one arc-minute of angular resolution on your retina is the average human being’s ability to resolve detail. Right now most of us are doing one arc-minute. If you’re a navy seal sharpshooter you’re sub, some people are as high as 1.5, 1.6 arc-minutes if your vision is getting worse, so it’s a bell curve. Once you get to that zone it is not about anything else other than can you drop as close as possible to one arc-minute on the retina at all distances and all sizes. But when you think about putting a screen in front of someone’s eye, you think pixels and you think 4K and 8K and that’s a completely different mentality so our thought was always how do we get the angular resolution in a completely different process onto the retina in more of a volumetric way. So it’s just a different mindset to get people away from that 2D screen thing into something else.
Hamilton: Ok. I expected [the comment] to be more about varifocal research you guys have been working on.
Abovitz: We probably have one of the biggest I would say IP portfolios in — I think of them as like ‘zone focused’, varifocal is another way to think of it — like the light field you’re wearing now is infinite. But our initial theory was the brain is not pulling in an infinite lightfield signal. The brain is actually chopping it up. And if you learn how the brain might be doing it — that was our starting theory. The human brain is not needing the fullness of that signal. So the physics of the pure lightfield is not what the human brain needs. So we have been chasing — what does the human brain need to be as physiologically accurate as possible? And that in itself is a very difficult job but a lot more solvable than trying to get the pure physics of a perfect analog light field which our belief is that the human brain never needs that like bat and cats and dogs they all take a different piece of that signal. So all we’re chasing is for ML1, “what is physiologically safe” and then as we keep getting better — like ML1.x, ML2, — how do we continue to improve it to where its like closer and closer to what our brain is really doing. So we’re sort of chasing the brain side, not the physics side. We actually do have a paper…
Hamilton: I still want those papers, you’ve been promising papers.
