What is Magic Leap really?

pulled from a comment I wrote

Magic Leap was banking on fiber scanning displays. The technology, however, turns out to be not ready yet, and likely not for at least 5 years.

Fiber Scanning Display in 2015. Image from kguttag.com.

How Fiber Scanning Displays work. Image from ASME.

So what are they using? We know this, based on all the leaks and info given out. They’re using a waveguide (like HoloLens, Lumus, Vuzix, Epson, etc.). They’re likely going to use LCOS (like HoloLens). They may use micro-OLEDs (like Epson) but probably LCOS for its low price, high brightness, and small size. Famed analyst Ming-Chi Kuo reports that Magic Leap will use Himax LCOS displays (likely an unreleased 1080p version).

Viewable area of the HoloLens waveguide. Darkened shield for reducing external light to enhance brightness of virtual images (waveguides are notorious for light waste). Image from Neogaf.

An issue with using waveguides is that they’ll only give a relatively low field of view. Lumus has a state-of-the-art 55 degree FOV waveguide. While the FOV will continue to grow, it will probably not be as large or larger than Meta’s FOV (which they claim to be 90 degrees) for a few years. It’ll be better than HoloLens (gen 1) which has only around 30 degrees but nothing mindblowing.

Field of view diagram. Image by Chris Grayson.

How are they going to differentiate? Multiple fields of focus. They plan to use multiple waveguides stacked on top of each other to enable that. However, that’ll make the waveguides significantly thicker than they already are. Also, if they want multiple focus depths at the same time, they’ll sacrifice frame rates. Most likely, they’ll use eye tracking to figure out what field depth to project at based on where you’re looking, to maximize frame rate.

Diagram of a Magic Leap multiple depth field patent. Image from kguttag.com.

For processing, they’ll split it out to two packs, in your pocket, or more likely strapped to your hip since heat is apparently such a big issue. Battery pack and processing pack. They’ll likely use a Snapdragon, or whichever is top of the line when they release. Not only will using ARM over x86 save on power but it will allow them to leverage the foundations of Android or whatever microkernel that Google creates in the future to further save power and have fast accurate spatial mapping.

The last time people wore computers on their hip. Image from Pocketnow.

All that processing is going to go to 3D spatial mapping of your environment and of your hand. HoloLens currently has the best algorithms for this and even a custom-built silicon chip to handle it. This will be an important component for any AR device. Magic Leap could make an equivalent or better system in theory, but Microsoft certainly has the lead with the best system and it will be difficult to pass them in the near-future if you’re just a startup.

So what does Magic Leap appear to be in the end? What makes them unique and competitive? Not all that much, it seems. If you tear away the veil of hype that Rony is a master of, they have no competitive advantage currently. They have no unique qualities or technologies that their competition does not have or cannot catch up to in short order. Microsoft is the leader in AR right now and is best positioned to lead even when Magic Leap finally releases version 1.

Rony wants to be a “Baby Apple”, but you can’t be like Apple if you release Yet Another Waveguide. The iPhone wasn’t a BlackBerry or a Treo. It was something entirely different that hadn’t yet been conceived (except for crude attempts by others that went silently into the night). Magic Leap will not succeed against a juggernaut like Microsoft if they release a modified HoloLens. Microsoft certainly knows that you can’t just release YAW, as they’ve now decided to skip version 2 and go straight to version 3 of HoloLens with far superior technology for a 2019–2020 release. I wouldn’t be surprised if Magic Leap wanted to do the same. Unfortunately for them, it may be too late to catch up to competitors’ traction by then, even if investors are willing to give them time to perfect their first release.

How could Magic Leap catch up? By releasing something amazing. But all indicators point to no significant competitive advantages and the FSD will likely not be production ready in only 2–3 years at such a disorganized startup as Magic Leap.

I could, however, imagine them releasing an equivalent competitor to HoloLens. They could gain some traction that way. I am not saying it’s impossible. But even if they miraculously did manage to get on their feet against the competitors they will have at release time, they face a far bigger threat.

Apple and Samsung will likely dominate consumer AR in the 2020’s, if not beyond. Why? It’s simple: they have traction, money, and technology. Let me expand on these.

1. AR is compute-intensive, and there is already a top-of-the-line computer in your pocket made by them. Magic Leap will not only have to sell you a display, but a computer made just for it also. That’s easily doubling the cost. Meanwhile, A/S (Apple and Samsung) will have already sold you a pocket computer. That means that a $999 AR device from Magic Leap turns into a $499 AR display from A/S. Even if Magic Leap changes to leveraging the phone you already have, they will not be able to overcome the native advantage A/S will have, and will never have access to the rich customer base Apple has because they’ll never be able to leverage iPhones with the control Apple exercises.

2. They have traction that Magic Leap will never be able to surpass. This relates to the first point in that an AR device will be for 5–10 years from now an accessory device rather than an iPhone replacement. It will be part of an ecosystem centered around your smartphone. There are two scenarios I envision here: one, Samsung, Microsoft, Magic Leap, and other smaller competitors release their products and iterate on the technology year after year. After a few years, Apple jumps in and sweeps the market with a functionally equivalent device that beats everyone else on the finer details and ecosystem (including rich customer base). This is essentially the Apple Watch Scenario. Two, Samsung, Microsoft, Magic Leap, and other smaller competitors release their products and iterate on the technology year after year. AR devices remain a niche and are not widely adopted. When the technology matures, after 5–10 years, Apple jumps in with a technologically superior device, and sweeps away all but their biggest competitors (Samsung, Microsoft, Google). Apple dominates the consumer market for a while, Samsung competes and eventually catches up, and Google works fast to release their competitor. This is essentially the iPhone Scenario. With the way Apple’s leadership has managed the company since Steve Jobs died, I would bet on something closer to an Apple Watch Scenario than an iPhone Scenario.

3. “If you see a stylus, they blew it.” Steve Jobs said something to that affect about the crude tablet technology that came before the iPad’s pure touchscreen (and also about the smartphone). Likewise, I would say “if you see wires, they blew it.” AR technology will never explode with the consumer if you have to wear a beltpack wired up to your headset, or even your phone wired to it. You might point out that wearing earbuds wired to your phone is perfectly normal, but we are all familiar with bending too far or bouncing too much and having our earbuds yanked out or falling out. No matter what, Apple’s ultimate vision is wireless, and I suspect that they would strongly prefer to release glasses that simply relay most of its non-immediate processing (everything but millisecond-crucial spatial tracking) to the iPhone wirelessly (and just load the rest of the device with batteries and display hardware). Competitors will come out with wired AR glasses first for the foreseeable future, but wireless (whether or not processing is self-contained or off-loaded) is the inevitable conclusion and whichever companies end up dominating the market will ultimately need to go in that direction.

Analyst Ming-Chi Kuo also said that Magic Leap is paying for a high-priced high frequency antenna, which could mean that their device would wirelessly transmit to a beltpack. This would definitely be better, however point 1 still stands. On the positive side it could be “just good enough” to gain a foothold in the early adopter market.

Steve Jobs explains design decisions for the iPad. Video from All Things Digital.

So what will become of Magic Leap? Most likely, they’ll compete for 5–10 years from now, and eventually they’ll be bought up by Google or Samsung or Snap (which could be a wildcard competitor). They’ll either be used as a jumping off platform if they have a successful future, or as a patent bundle if they don’t do so well. I could imagine that ten years from now their fiber scanning display could be mature and ready for production. I could also imagine that it could be a dead end.

A wildcard player could come from Snap (formerly Snapchat). Facebook is also a possible contender. Video by The Verge.

In the end, Magic Leap is a hype-train with no significant advantages over its much better equipped likely competitors. On that note I could also see Vuzix being acquired for their display technology in the nearish future. Magic Leap will likely be acquired, also in the nearish future, most likely by Google/Alphabet (one of Magic Leap’s biggest investors, with Google CEO (not Alphabet) having a board seat) or Alibaba, but also possibly Samsung, Microsoft, Apple, Facebook, or other potential players.

This is, of course, a Western-centric point of view. Alibaba is part of a $700 million+ funding round in Magic Leap and Magic Leap has released a video showing off a collaboration with Alibaba’s Amazon/Ebay-like platform Taobao which is one of the largest in China (and the world).

Taobao shopping demo on Magic Leap. Image from Magic Leap.

AR is perfectly suited for Chinese society. China has incredible integration with mobile technology that America and Europe does not. Their entire lives are run through mega-apps like WeChat (owned by another mega-corp, Tencent, which competes with Alibaba). The paradigm shift with AR is that it turns mobile technology from an inside-out device that you consume from into an outside-in world that we live in (which could be called Mixed Reality, or MR). China is suited for this with their society’s tight integration with mobile technology on a level which is simply not seen in Western society. They live in a world built on mobile while we live in a world where mobile simply augments our current world.

Alibaba, or even Tencent, could easily acquire Magic Leap in 5 years and use them to jump off AR in China. Tencent has invested in Meta so they are certainly looking into AR as well.

Man wearing Meta 2. Image from UploadVR.

To bring this home, Magic Leap will probably be acquired, after 5–10 years, like Rony’s last company. They’ll either be bought for their patents or for their platform and technology. They possess no current competitive advantage besides being one of the few current players preparing to enter the AR market in the near future. Their fiber scanning display may eventually mature into a technological advantage, but it could also simply be a dead end (although with the release recently of a supposed virtual retinal display for people who can’t see well in Japan, I wouldn’t count it out yet). Magic Leap could have a competitive edge for a little while with its content partnerships (ILM, Warner Bros, Weta, etc.). But I would bet my money on them either being acquired or simply failing and being sold off to investors.

Check out where I got a lot of knowledge about Magic Leap technology: http://www.kguttag.com/