Shower thoughts on the Magic Leap One

We are in the days of the Magic Leap hype: developers that have bought it are posting videos everywhere and every one of us is trying to understand if the device is valuable or not. In these days I’ve read lots of reviews, tweets, and comments from the community and so I’ve now a personal idea on this device and its company. So, here you are my editorial regarding Magic Leap.

First of all, I want to talk about the device: from all reviews, it appears as a solid HoloLens v2. Some opinions are more enthusiastic, others are colder, but I’ve read nowhere that the device is not valuable. This means that we are talking about a good augmented reality headset. I said that it is like a HoloLens v2 because it seems better than HoloLens almost in everything, like for instance:

• The graphics appear crisper;
• The number of allowed polygons is superior (if I remember well, we are around 50-100K polygons in the scene vs the 20-30K of HoloLens);
• The environment tracking is slightly better (apart from black and reflective surfaces, that are tracked really bad);
• There is hand tracking and eye tracking;
• The virtual objects are perceived at different depths and this increases the realism a lot;
• The FOV is 45% bigger than HoloLens and has also a nice blending feature that reduces the effect of having a see-through window.

There are other things that seem worse than HoloLens, though. For instance, tracking stability: on Magic Leap, the “holograms” appear as trembling a bit, while on HoloLens, they’re damn fixed. When I’ve tried HoloLens, I was really impressed by this, because I know how it is hard to make no-jittery augmented reality and Microsoft has really done a terrific job in doing this. Then there is the HPU: IMHO, having a separate computational unit completely dedicated to performing 6DOF tracking relieves a lot the work of CPU and GPU. So, even if the computational unit of the Magic Leap One is far superior to the one of HoloLens, part of it has to continously work to perform the tracking and the environment scanning and this degrades the performances. Then there are various little glitches and bugs of the device, that are pretty normal for what Magic Leap One actually is, that is a dev kit. (P.S. for an interesting comparison between Magic Leap and HoloLens, read this post)

Pros and Cons, as in every AR headset out there. For instance, the Meta 2 has a huge FOV if compared with HoloLens and Magic Leap, but it is a tethered device and this means that you are bound to stay at your desk to use it. The Magic Leap appears so a good AR headset, that improves the old HoloLens in many parts. There are also features into which the Magic Leap is truly an innovator with respect to the competition. One of this is design: while it isn’t a device that can be worn in the streets, it is surely less nerd-looking than HoloLens, it is more classy. This can make the Magic Leap One appeal to new categories of customers, like the ones in fashion and design, that wouldn’t have liked the geeky HoloLens. And then there is comfort: it is not only very light, but there is also a big attention to comfort, with lots of options to customize it so to fit your face (nose pads, cushions, etc…) so that to improve the user experience. There is also the UX: there are few Magic Leap experiences available, but all of them are damn cool to be seen. All the videos shared on the social media are making me want to try it. Even the menu is amazing, much more beautiful to be seen than HoloLens one. I can also talk positively about the shipping experience: they launched the product and THE DAY AFTER there were already developers having it at home. That’s impressive.

And all of this has been done by a little startup somewhere in Florida, not by a major company. In a standard case, we would be all talking about a very interesting product created by a cool innovative startup. Reviews will be all positive. But not in this case.

The reason? Well, we all know that. The hype (and bullshit) machine. In all these years, Magic Leap has created an insane hype towards its company, with fake videos (like the fake CG one where they played a shooting game in the office), super-secretive info, nonsense TED talks (really, WTF) and bold claims. I don’t forget: the CEO told a lot of times that Magic Leap was able to make virtual objects appear as real. He said that they were far better than the competition, he even said that the competitors’ optical systems were bad for the eyes of the users. He promised a device for consumers, that would have been far ahead of the competition, that would have taken the future into the hands of people.

But this, of course, hasn’t turned out to be true. Because… well, because the technology isn’t here yet. It was all to create hype. And no one forgets these promises, so journalists when reviewing the device haven’t said: “It is a cool AR device, even better than HoloLens”. They are saying “Yes, nice AR, better than the competition… but it is NOT as the real world, it is NOT what this company has promised us“. Reviews are all comparing the device to the promises, resulting quite negative and skeptical towards it. This has hurt the company launch, because no one of us really felt compelled to own the device after having read these reviews.

Journalists highlight that Magic Leap is better than HoloLens, but HoloLens is a very old device. We are comparing a headset of today with one that has been revealed three and a half years ago (I don’t know why some journalists say that HL is of two years ago, it has been revealed in January 2015). Three and a half years ago is an enormous amount of time in computer science.

At the beginning of 2017, Microsoft stopped thinking about releasing a HoloLens v2, because it had no real competitors and the AR market was in any case small, so the effort of releasing a new version had little sense. We don’t know how this HoloLens v2 would have been, but most probably would have improved a lot wrt the previous one, so maybe it could have been on par with the Magic Leap One regarding FOV, processing power and environment reconstruction capabilities. Regarding design and comfort, instead, I think it would have been very unsexy anyway because HL is dedicated to industrial settings, where the esthetical appearance is useless.

Regarding the software, now everyone is praising the feature of the Magic Leap’s Helio browser that lets you take 3D stuff from web pages and put it in your physical room… but I clearly remember having seen such a similar thing in a Microsoft event streaming when I was still in Immotionar (so, more than one year and a half ago). I haven’t found the link, so I can’t show you the proof, but this means that Microsoft already has such a functionality behind the curtains. Then everyone loves the fact that with Magic Leap you can make the real world occlude virtual elements. Wow… but this was already somewhat possible with HoloLens: you just had to take the mesh of the environment and assign it an occluding transparent material… I don’t know why everyone is forgetting this in these days.

Personally, I think that the HoloLens 3 will be better than Magic Leap One. Microsoft has more money and more expertise than Magic Leap (it has made the Kinect…) and works actively in sectors like AI (Microsoft Cognitive Services), cloud (Azure), body tracking (Kinect), etc… since years. It has already teased that the new HoloLens will have AI features, hand tracking, and a bigger FOV. It has the feedbacks of lots of customers that already have HoloLens (more than 50000, according to official numbers), so will be able to create a device that fits more the requirements of users. Magic Leap, on its side, has just launched the device in the wild.

Returning to the hype machine, I think that it has been the evil side of Magic Leap. I’ve hated and I still hate it these days. But I can also see that it has been the thing that is making Magic Leap successful. There’s a saying that says “There’s no such thing as a bad publicity”. All the fluff created by Magic Leap and the resulting backfire has given the device a lot of visibility. This, together with the secrecy, has made people so curious about the device that as soon as there was a little news on Magic Leap, it was spammed everywhere on the social media. We all hated Magic Leap hype machine, but we all wanted to know more about it. The result has been that when the device has been released, 1000 people have bought it almost immediately, even if it cost $2300. Absolutely not bad.

MagicLeap had around ~1000 orders placed on its first day of sales so far (with limited distribution in the U.S.), each one containing up to 3 devices.

Revenue today was probably somewhere in between 2.25 and 3 million. What do you think of the numbers?

— Lucas Rizzotto (@_LucasRizzotto) August 9, 2018

And they have not only bought it, they have started experimenting with it, posting GIFs on social media, giving more popularity to the device. The device can be bought only in six cities and this, in my opinion, has the following advantages:

• The startup has been able to execute fast, shipping the headsets in no time;
• The shipping management has been easy;
• Giving the headset to a few people, it is able to get feedback and fix problems before launching worldwide;
• The company will have longer visibility: as soon as the device will be released in more cities, new people will start posting photos and videos of the Magic Leap, giving a continuous visibility stream to this company.
  It has been a genius idea.

Every new platform needs desperately developers. And thanks to this hype, Magic Leap has obtained developers that since day one are working with the device, are willing to create apps for it. Think about other AR devices out there: how many people do you know that are developing for an AR headset that is not HoloLens? There is an interesting project out there called Aryzon that offers you a cheap AR headset for less than $40 but this startup struggles to find many interested developers. Magic Leap, instead, thanks to the crazy hype it generated, has been able to attract the first devs easily. So in the end, I have to say that all that hype has been the right choice, because it made they obtain the result that they wanted.

So, today, Magic Leap wins. But we have to see how it will behave in the long term. Excitements of the first days finish easily and now that the headset has been revealed, there’s no more magic around it. They’ll have to find a new way to keep it cool and compelling for us developers. Then, I think that sooner or later they’ll have to pay for all the annoying stuff they have said in these years: there are various influencers out there that don’t love Magic Leap for this and they are continuing posting skeptic stuff on social media. You know that I myself don’t love this company for exactly this reason and I have made a little video mocking it doing a Magic Leap emulator for the Vive Focus.

Lots of people like me, that can’t afford the Magic Leap now, will wait for the next HoloLens to see which device is better to buy. Maybe another company would have caused people to buy the device to support its efforts, but this is not the case.

And Magic Leap needs support considering that has now to face the competition of a giant like Microsoft, that for sure is going to answer with a great HoloLens v3. And in the future there will be other big players like Apple, that already plans to revolutionize augmented reality, Google that for sure will want to revenge the Google Glass and Facebook, that is already experimenting with AR inside Facebook Reality Labs (Abrash has confirmed this). All big companies, with deep pockets. Plus we don’t have to forget that all these companies are already somewhat competing with Magic Leap with mobile AR. ARCore, ARKit, and Facebook Camera are augmented reality frameworks that already work well and that are improving fast: multiplayer, environment reconstruction, virtual elements occlusion and persistent consistent experiences among different players of the world are coming (if you follow companies like 6D.ai, you know that cool stuff are coming for mobile AR). I’m not a huge fan of mobile AR, because I think that holding a phone in the hand to live augmented reality is uncomfortable, but it somewhat works. And it has a HUGE user base of hundreds of millions of people, that as soon as will find AR useful or cool, will use it everyday as they use Snapchat filters.

Yes, there is the content and market problem for Magic Leap. HoloLens is a device for enterprise and has established a great reputation in the industry sector. Microsoft has also released some software that is fundamental for big factories. Rony Abovitz instead continues saying that Magic Leap is not for enterprise, but for “creators”… it is also for the kid in the garage that wants to create, he says (around Abovitz’s house there are a lot of rich kids, I guess). No one at Magic Leap says “dev kit”, because this is not a device for techies, but for creators. Sorry Rony, but all the people that I am seeing using the device are actually developers. Maybe the message was not clear.

The problem is that no consumer would ever buy a super-expensive dev kit toy: it is too bulky, too difficult to be used, there are no useful apps and it is too expensive. So… what is the strategy of the company? I am a developer and I know that if I’m developing a software for HoloLens it is mostly for some enterprise use. What about Magic Leap? Why should I create something there? Who is going to buy my product? Am I just creating it so that ML can have a more compelling device to sell? Well, sorry, in this case, I am not interested in offering my free help for the above reasons. What market am I expected to have for my apps? No consumer will buy the device… so am I developing for the other developers out there? And… for how long the people are using this device? Are you sure that it won’t suffer from the Gear VR effect, so it is cool for some days and then people will leave it on the shelf? Am I developing because I bet that this platform will win against Microsoft, Apple, Google, and Facebook? Well, the stakes are not that high, unless some big corporation is going to buy the company, as it has happened with Oculus. These things are all unknown at the moment and no one is giving us an answer.

The sooner Magic Leap realizes that its device at the moment CAN’T BE tailored at the average consumer the better is for its ecosystem. It has to find a direction, a target market where the device is compelling and stick there until times will be ready for mainstream adoption. If factories are not sexy for them, they can try with art, interior design and fashion for instance. In this moment they could sell themselves saying that their product is better than HoloLens and costs less, and also has warranty, so this is the moment that ML should push in every industry it can. This is the moment when they have to find their sweet spot.

I’m not buying a Magic Leap because I don’t live in the US and I don’t have that money to invest in. I would experiment with it at the moment only for curiosity, to have more likes on social media (Magic Leap’s videos are getting lots of likes and retweets) and to appear as a cutting-edge consultant with customers. Honestly, since from the reviews it doesn’t appear as a game-changing device, but just an incremental step forward, I will wait next year to see how will be the next HoloLens and if Facebook and Apple will make their move. And especially to see if mobile AR will take foot. Then, I will decide what to buy. In the meantime, I will continue having fun with my Vive Focus.

But I’m happy that Magic Leap has released its device, because after all the investments and attention it received, if it failed, it would have hurted the trust in the whole AR ecosystem. And then this way, finally there is competition in the AR market and we have already seen in VR how competition between Oculus and Vive has been beneficial to lower prices and obtain cool features.

In the end, I can say that I find interesting the Magic Leap hardware, but I’m worried about what there is around it. After all these years of wait and the final release of the first product, now Magic Leap has to prove that it can be a reliable company with a compelling software platform. I think that the great thing of Microsoft is that it gave a clear direction to its HoloLens, while Magic Leap still thinks about the dream of mainstream adoption, a dream that maybe will come true in 5-10 years. I hope for them that they will be able to survive until that moment.

(Header image by Magic Leap)