With $590M in more funding, Magic Leap looks to the future | interview

Magic Leap announced in a filing that Saudi Arabia’s Public Investment Fund has poured another $590 million in debt funding into the mixed reality headset maker.

That is happening just as Apple prepares to go head-to-head with Magic Leap in the mixed-reality goggles market with the February 2 launch of the Apple Vision Pro. I talked to Ross Rosenberg, the new CEO of Magic Leap, and Daniel Diez, CTO, in an interview last week at CES 2024.

While Apple’s device will be tough to compete with, Magic Leap has a slight pricing edge, with a price of $3,300 versus Apple’s $3,500 device. But it is interesting that Saudi Arabia’s Public Investment Fund, which is already the majority owner of Magic Leap, wants to double down on the investment. (The investment date was in November). To date, Florida-based Magic Leap has raised $4.5 billion.

Magic Leap’s founder, Rony Abovitz, stepped down from the CEO job in 2020. Peggy Johnson succeeded him and then she was recently replaced by Rosenberg, a former Bain Capital executive. I spoke with Rosenburg, who talked about the vision for solving customer problems, and Diez, who has been there for some time, answered questions related relating to Magic Leap’s tech and history.

“We have these hype cycles, and that’s ridiculous. What’s interesting is the in-between times, when people actually put their heads down and figure out what they’re going to do,” Diez said.

Here’s an edited transcript of our interview.

Daniel Diez: We’re not on the floor this year. We’re here in these private suites now because we’re moving into more of a go-to-market part of our journey. It’s less about the broad awareness and more about helping customers connect with specific use cases and understanding how they’re using these technologies. That’s what we’re spending our time on here at CES. And of course it’s Ross’s first CES as CEO of Magic Leap, so it’s been good for him to see our customers and our partners.

Ross Rosenberg: It’s been a great opportunity to see all the progress being made in XR and our pioneer status in that. It’s a pleasure to be here.

TechForgePulse: Is anything changing in the time you’ve been in charge now?

Rosenberg: I’ve been in the role for about two months now. I was excited to join Magic Leap because of its technology lead, and because of the use cases for customers. Solving real problems for customers has made a lot of progress in that way. That’s what brought me here. We’ve been thinking a lot about the third chapter in Magic Leap’s history. Chapter one was about invention and the breakthroughs that were created under Rony. Chapter two was about productizing those inventions. We now have Magic Leap Two. Chapter three is about commercial adoption of that technology, which goes to customer use cases. It goes to applications. It goes to content and all the things we can help the industry coalesce around an ecosystem.

TechForgePulse: Is there any one product category or industry category that’s doing the best for you?

Rosenberg: The way we think about it, the use cases where you can do the kind of collaboration and interaction with digital information, and in a way where there’s simply no substitute from a 2D device. An example of that would be in health care for surgery, complex surgical planning, whether it’s brain or spinal surgery. The surgical preparation, and even the surgery itself that needs to happen. There’s simply no way to do that level of planning and get the accuracy ahead of time without doing it in a collaborative and spatial way. Same thing in the military space. We think of those as very high value use cases for this device. We’re seeing a lot of traction there.

Diez: We’re looking at industrial settings, training for complex procedures that require large scale digital twins and sub-millimeter tracking and overlay of that content. That’s another area where optical see-through, and specifically Magic Leap’s technology, serves a portion of the market that can’t be served by any competitive spatial computing.

TechForgePulse: Sub-millimeter, is that a good way to convey how precise it is? What does that refer to?

Diez: The digital overlay, when you want to overlay content and have it stick over the physical object, that’s where you are. If you want to be able to modify that–a great example is this company called Medical Eyesight. They’re out of the U.K., and they’ve been regulatory approved for surgery. They do mechanical thrombectomy, physically removing blood clots from the brain. That’s faster than waiting for medication to do it, and the faster you do it, the less likely you are to have extensive brain damage and lose function.

They’ve been able to do that now through guidance with Magic Leap. The vessels they’re in with the catheter are re-created in front of them, and they’re able to guide through the human body with that 3D model. Because it’s so precise, they’re now able to reach areas of the brain that were impossible to reach with the previous methodology of using a 2D screen for guidance. That level of accuracy is enabling these use cases that require high precision. They’re having a major impact on patient outcomes.

TechForgePulse: How do you feel about your defensible space relative to Apple coming in?

Rosenberg: When you have the desire to not just display information, and not just even locate information, but to interact with that information at that level of precision, there’s no one else that can do what we do. There’s a big moat around the company for those types of use cases. What Apple has shown, as I’m sure you’ve seen, is much more about watching a streaming movie or a Facetime call, where you don’t need nearly as much accuracy. That’s a very different space.

Diez: The other big difference is that if the task you’re doing requires you to be in motion as the user, or if things in your environment are in motion, the lag of re-rendering every time an object moves or you move and putting it on the screens is a deal killer for something like military training applications, police, first responders, active shooter training, or physicians who want to see things immediately. We’ve carved out a defensible position in that this is one of the few headsets, maybe one of one sometimes, that can do what Magic Leap does.

TechForgePulse: Has the price stayed the same, or have you reacted in some way to where Apple is at?

Diez: The pricing has remained consistent since we launched.

Rosenberg: We’ll look at it through the lens of the customer and the value we deliver. In some of these very high value use cases, either because the device has been certified for medical use, or because it’s delivering so much incremental value for our partners, the pricing that a consumer might evaluate is somewhat irrelevant to the way people think about the ROI from this technology.

TechForgePulse: Are the digital twin installations turning out to be good for the company as well? Siemens had a very big keynote here on the industrial metaverse. Is that something you’re part of?

Diez: Siemens is a good partner. We’re working with a number of clients using their digital twins with different use cases. It’s a big part of what we’re doing and where we deliver value.

TechForgePulse: Is that potentially something where you have a chance to sell in thousands of products, one of the big efforts or big companies?

Diez: That’s the road map, moving this from POCs to larger-scale deployments. As more and more providers like Siemens, like Cisco jump into this game and provide that halo of having a hardened enterprise solution, combined with the cutting-edge technology, I think that’s definitely the recipe for adoption at scale.

TechForgePulse: What is the road map for where you go from here?

Rosenberg: We have a road map based on expanding the technology we have. Expanding the field of view, expanding the immersiveness, making the device more mobile in some ways, less tethered. All of those things are on our road map. As you extend beyond that time period, you think about ways to further reduce the size and weight of the device. Eventually you get to something that feels more like glasses. We think about all those things as we design our road map.

TechForgePulse: Do you use Qualcomm chips, or something else?

Diez: Right now the chip is AMD. The compute power necessary for this device is large. You’re looking at a Macbook Pro level of computing. That requires the semi-custom chip that we designed with AMD. It’s very powerful. Something lighter weight, maybe meant more for a mobile device, would not be able to do what we need this to do.

TechForgePulse: With semiconductor strategies, sometimes I think companies will opt for a platform that somebody else can invest in more, and then you get the benefit of that. As opposed to having to do the work of custom design. Even working with another party, as in this case, if some other entity can do some of the work–that’s always Intel’s pitch. “We’ll take this off your hands.”

Diez: At the time, when this was in development, the partnership with AMD was critical, their ability to design with us and meet specific requirements. The chip that they’ve created is semi-custom. It’s available to the market. It’s not completely exclusive to Magic Leap. AMD’s bet is that they want to continue to develop these chips so they can supply this market. As the market progresses, you’ll absolutely see more chip manufacturers begin to develop chips that will have this level of compute, and they’ll try to truly compete for what will be a very large hardware market for AR and spatial computing.

TechForgePulse: At that point, is there some other value that you’ll focus on?

Rosenberg: There’s a few things. The optics technology we have that allows the immersive experience, no one else in the world has it. The broad field of view, and it’s only going to get bigger as we execute our road map, is something only we can do. We don’t necessarily think of that as distinct today from the level of compute that’s necessary, because we also have software that allows you to perceive the world. It allows digital information to stick to the real world as you see it. That’s very important. As other people get better at that, although they haven’t yet, then it could allow you to step back if you wanted to. But today–our customers require us to do all of that.

Diez: As more componentry becomes available and it gets more advanced, Magic Leap will be very happy not to have to custom create things ourselves. Where we would focus our time is where the majority of our IP and value lies, which is in this optical stack. That’s next to impossible to re-create.

TechForgePulse: Are advances in AI bringing something new to the table for you?

Rosenberg: There’s a history of AI in the device itself, just to recognize what’s in a room with you. We’re moving the ball forward pretty significantly, because if you think about an AI model, a generative model, being able to light a certain area of the room and recognize everything that’s there, that’s a rich source of data that we provide. At the same time, all the behavior change that’s occurred with generative AI–people want their output first in text, second in images, third in 2D video, and eventually they’ll need that same output to the same questions in a spatial way.

TechForgePulse: How many people does Magic Leap have now? Has that fluctuated much?

Diez: We’re over 1,000. We’re probably over 1,100 at this moment in time? I don’t think that has fluctuated for a bit of time. We have a number of positions open right now. We’re definitely growing the team as necessary.

TechForgePulse: “Metaverse” isn’t as popular a word today as it was for a while. What’s your perspective on that?

Rosenberg: Customers have become much more sophisticated. They can be much more discerning. They don’t necessarily need those kinds of broad-based terms. Again, am I going to display the information? Am I going to locate it? Am I going to interact with it? They really want to know what kind of spatial 3D experience they can have. How does it integrate with data that sits inside my company? They end up being very specific and precise business process and workflow kinds of questions. Metaverse did a good job of driving curiosity, but many customers have moved beyond that at this point.

Diez: The way it was defined early on, it also became incredibly narrow. Because it was one company leading the conversation at the time, it became very much about artificial digital environments that remove you from the physical world. It became much more of a VR conversation. People were fascinated by that, which is great. Any interest in spatial computing is a great thing. But it became too narrow a view of what the metaverse is.

It’s pretty commonly accepted or agreed upon that the true capabilities of the metaverse will come to life when it’s a fabric built of digital experiences ingrained or embedded into the physical world. That’s where the metaverse becomes really interesting. You throw around terms like “industrial metaverse,” talking about digital twins of factory floors or specific equipment, those aspects of the metaverse are what customers are most interested in right now. That’s where the value is derived. But if we do this right, the metaverse will be not about this idea of world escapism and going into digital environments. It’ll be the thing that enables us to be even more fully immersed in the physical world, doing things that we weren’t capable of doing before.

Rosenberg: Most of the customers we talk to are trying to get something done. They’re trying to repair something, train someone, design something. Those are the words they use. They don’t start with, “Hey, tell me about your metaverse.”

Diez: It’s a backbone thing that they don’t really talk about anymore. It’s just understood that there are aspects to what we think the metaverse is that they know are necessary for their applications.

TechForgePulse: It’s an interesting side business for Sony to come up with their own headset now, to design metaverse experiences. It doesn’t seem like their core business.

Diez: PlayStation VR is doing well for them, I think. It’s an interesting area to explore. VR and passthrough is an ideal technology for gaming and entertainment purposes. It feels appropriate for them. I can see why they’re exploring that. We’ll see where they go next.

We have these hype cycles, and that’s ridiculous. What’s interesting is the in-between times, when people actually put their heads down and figure out what they’re going to do. Inevitably that leads to another hype cycle and people get disappointed again, because it doesn’t do what they want to, but then you go back and do the work again.