Magic Leap is looking for more money

So it appears Magic Leap is looking for even more money. This would be on top of the ~$1.4 billion already thrown at the company.  That amount is already unprecedented so talk of further investment is nothing short of astonishing. The thing that got me interested in the company in the first place was the ludicrous amount of investment they had raised from such interesting and respected sources. Piling more on top of this is just wild. 

So it appears Magic Leap is looking for even more money. This would be on top of the ~$1.4 billion already thrown at the company.  That amount is already unprecedented so talk of further investment is nothing short of astonishing. The thing that got me interested in the company in the first place was the ludicrous amount of investment they had raised from such interesting and respected sources. Piling more on top of this is just wild. 

If they are able to raise this money, this is great news for Magic Leap. They have already gone through a small PR firestorm last year when an article from The Information (paywall) cast doubt on some of the more ambitious speculation surrounding the company. Many people and even publications have held this up as evidence that Magic Leap is in big trouble.  This largely stems from the sort of cynicism and entitlement that is rife across the internet and does not accurately reflect the article in The Information. I have already spoken about my distaste for this towards the end of this article so I won’t bore you with it again. Regardless of why, and regardless of its baselessness, public sentiment is against the company. Given that climate and the amount already sunk into the company, a raise by Magic Leap at this point is all the more impressive and promising for their inevitable release.

Magic Leap reportedly had a demo to their board (who, as a reminder, are their investors) in February. While we don’t know what happened in that meeting, we can only assume it went well if they are willing to up more investment. You have to think the likes of Google and Alibaba would take a particularly critical eye at this stage and given the public pressure on the company.  Passing this tests speaks volumes. 

Why do they need more money?

More money means that Magic Leap must be doing something right but they would only do another raise if they needed more money.  What have they spent $1.4 billion on? That is a massive sum. Karl Guttag has rightly pointed out that Avegant has been able to develop similar technology to Magic Leap with significantly less investment. I don’t want to minimize what Avegant has accomplished as it sounds impressive but it is not a product. It is a prototype. Presumably it is a prototype similar to what Magic Leap has likely had for some time now, perhaps as far back as 2015.

Building a product, not a prototype, at scale is hard. And expensive. It is particular hard and expensive when you decide to build not only the product but the factory that makes the product. This is almost unheard of. Not even Apple exerts this sort of control over its supply chain. It becomes even harder and more expensive when you spearhead a large portion of the content creation in house. Only Nintendo can claim to do something similar. Meanwhile, you are still maintaining a massive R&D effort to continue to build on the first iteration of the product and think towards the future.  These sorts of R&D labs are what Google and Microsoft are known for. While all of this is ongoing, you have to engage with other companies to form a developer ecosystem (which appears to be happening behind closed doors rather than in the open) so that the content you produce is supplemented by 3rd parties. This is all expensive.

Magic Leap is trying to be the best of Apple, Google, Microsoft and Nintendo all in one package and they are trying to do it pre-revenue. So yes, they need more money. And yes, what Avegant did is impressive but that part of the product is the “cheap” part. Don’t confuse this with the easy part. It still requires a monumental engineering effort. But the expensive bit, the bit Magic Leap is currently in the middle of, is making that engineering effort into a compelling product people want to actually buy. This is why Magic Leap has spent its money and this is why Magic Leap is raising. It is damn fascinating that it looks like they might be getting what they ask for.

Mixed Reality mapping of the world will change the world

We all use Google Maps. It is arguably the most useful part of a smartphone. Being lost is, to a degree, a thing of the past. It is one of the most important innovations in recent memory and is perhaps the ‘killer feature’ of a smartphone. If Mixed Reality becomes popular, and we stretch our privacy a bit, then we are about to see another massive shift in mapping and how we see the world in general.

We all use Google Maps. It is arguably the most useful part of a smartphone. Being lost is, to a degree, a thing of the past. It is one of the most important innovations in recent memory and is perhaps the ‘killer feature’ of a smartphone. Over the years, it has been improved and Google, among others, have been able to more accurately map and model the world. While it is nice to take a moment to appreciate the importance of maps, I bring this up to show how far we have come with mapping technology. And then I want to say that what is coming is going to blow that out of the water. 

What does Mixed Reality have to do with mapping?

On the surface, a Mixed Reality project doesn’t seem like it has that much in common with mapping. Of course there will be mapping products built using the tools of mixed reality. Some of the first examples of this already exist as seen with a Tango device. While a this makes for a compelling product and has the potential to be the main way in which we experience maps or navigation, it isn’t the revolution I want to talk about.

Passable World Models

Something that Magic Leap has talked about since the first patents started to come out, is the idea of a Passable World Model. A Passable World Model is just what it sounds like, a model of the world that we are able to pass to others. When you want to display something in mixed reality you need to map out (or model) the space in high detail so as to know how to position digital images as if they are in that space. That mapping is often referred to as SLAM and it produces a set of data that allows a device to know its relative position in an area. It is needed on the device but there is nothing to say it could not be shared with another user. This means that the other users device does not have to go through the effort of mapping an area. It already has the data. Someone else has passed them that model of the world. This frees up processing for the task of displaying mixed reality content. 

While this technique is useful for mixed reality, its application goes beyond just that. We are talking about mapping spaces in extreme detail. The data could be used to create an accurate and relatively high fidelity 3D model of the space around you. Imagine a fleet of millions of users mapping out everywhere they go and uploading those maps to share with others. This information can potentially be used to reconstruct a virtual version of the world. Think of a video game but the world you inhabit is an accurate representation of the real one. Google Maps, and Google Earth, already do this to a degree but street view is clunky and 3D maps are fairly low fidelity when you get up close. Further those aspects of the product, while cool and interesting, are less useful than the more basic mapping functions we use every day. People don’t use a mapping app because it has nicer 3D maps. So while this new development would be cool, why would it be so revolutionary?

 Google Earths 3D looks really good.... 
Google Earths 3D looks really good…. 
 As long as you don't get too close
As long as you don’t get too close

Near Real Time 

If Magic Leap becomes popular, or any other device that will utilize SLAM mapping, then millions of people will be using this technology all the time, everywhere. A google maps like product could be constructed that uses that data to update maps in near real time.  Want to know if parking is available? If a place is particularly busy? If something is open or not? What the weather is like? Or just what the lay of the land looks like right now.

You could build a grand theft auto style game but use a real city in real time that potentially uses Mixed Reality to interact with people in the actual physical locations. Think pokemon go but with digital characters controlled by people at home mixed into the real world.  All these things might be possible.

The potential of this data is mind boggling. You could join a party from the other side of the world. You could work remotely but it really feel like you are in the office and going to lunch with your colleagues. If you take it to an extreme, something like this could slow urbanization. If being remote is even a half decent facsimile of being in a place in person then why wouldn’t I live in a cheaper area that is potentially more beautiful than a smoggy city. This is obviously a far off vision and the reasons for urbanization are many but one reason is certainly centralization for work. If people are just as effective working remotely as they are in person then that reason might dry up to a degree.  

Privacy

I’ve deliberately avoided talking about the privacy implications here. They are important but I think they are also so obvious that it almost isn’t worth talking about yet. I don’t think anyone imagining this would not come to the same conclusions about potential privacy violations with this sort of technology. Simple measures can be put into place. Having to opt in non public locations. Blurring faces as street view does. Perhaps it is possible to detect and omit people entirely from the mapping, perhaps you can even do that client side. We could do the same with cars.  Perhaps replace them with avatars so you can’t recognize the vehicle. It is a conversation to have but I don’t think it is a particularly interesting one at this moment in time. We keep changing the bar of what we think privacy should be so by the time we actually are able to build a product like this, our moral position on privacy may be vastly different to what it is now. The rate things are going today, perhaps no one will care about these sorts of privacy violations in the future. This may or may not be a good thing but it is impossible to deny the current trajectory of the last 25 years.