When Microsoft first debuted HoloLens two years ago, it wasn’t clear if the company was working on a niche research concept that would find little life outside of tech demos, or if the company intended to push augmented reality as the Next Big Thing in the consumer market. HoloLens’ first release was strictly intended for developers and early deployment in enterprise settings, with its $3,000 price tag, lack of a warranty, and limited application support. In a recent interview, however, Alex Kipman, the inventor of HoloLens, said Microsoft has bigger plans for its platform over the long term.
Kipman introduced HoloLens to the world in 2015, and was previously admitted to Microsoft’s Hall of Legends in 2011 for his Kinect motion controller work. Kipman said HoloLens is much more than just a headset — it’s a set of software applications, an operating system, and an AI that picks up information about the world around you and then makes decisions about how that information needs to be presented to you, the viewer.
This touches on one of the biggest differences between AR and VR — one that has nothing to do with headsets, battery life, or application support. Virtual reality offers a different way to experience a game or application, but it doesn’t try to change the computer-user interaction paradigm. The additional processing power required for virtual reality is dedicated to keeping frame rates suitably high, tracking your position in the real world, and translating movements you make back into corresponding movements within a game engine.
HoloLens-HPU-die-plot
Microsoft’s custom HPU silicon. Image by EETimes
In augmented or mixed reality, by contrast, the entire point is to analyze and augment what the user is doing by making contextually appropriate options, displays, and information available when the user wants — not before, not after. Tying this information together means gathering data from many sources, analyzing user behavior patterns, and creating the operating system that ties it all together in the first place. Kipman said in the interview that Microsoft has been prepping this entire system rather than just components of it, but that bringing the product to the mass market takes a great deal of careful preparation.
Kipman said Microsoft “Abso-freakin-lutely” (emphasis original) has plans to debut a consumer version of HoloLens, but that simply asking about a ship-date is the wrong way to approach the topic.
The better question and the better way to answer it is, at what point is this thing going to be under $1,000? Because I can say it’s a consumer product tomorrow because I can remove the dev kit thing, [but] the $3,000 thing is going to get in the way of it becoming a mass market consumer product. You have to reduce the price point until it’s affordable to the majority of the populous of Earth, which will be under a $1,000 and then some to get there. Roadmaps for both of those things exist today, but I’m not going to announce or talk about it today.
Kipman also notes that he was a fan of Pokemon Go, saying: “The thing that I was fascinated by is the actual location database they provided and they curated to be able to bring that experience to market. And it was a magical experience that shows you the power of location, the power of spaces, the power of [overlaying] digital assets over the real world, even if it is a little thing. And it shows you that as soon as the content goes over the real world you have people running in parks in the middle of the night for a purpose. It’s exciting.”
It is exciting, but I’m not convinced Microsoft is the right company to bring the hardware to market.

Can Microsoft pull a consumer HoloLens off?

The problems HoloLens could run into aren’t all that different from the issues Kinect faced. Then, as now, Microsoft’s skunkworks project had invented an intriguing alternate method of interacting with a computer. Kinect has been used to translate sign language into text, as a way to build cost-effective three-dimensional maps robots could use to navigate their surroundings, and to even build robots that “know” when you want a beer. Kinect and Kinect 2.0 are actually pretty cool — once you divorce them from gaming, where they mostly tanked.
There’s no doubt there are environments and use cases for HoloLens, and Microsoft’s decision to focus on the enterprise for now indicates it was aware of the danger of pushing this technology into consumer markets first. At the same time, however, all technologies need use cases. Wearables have broadly failed to gain much traction because potential consumers haven’t seen much benefit to owning them.
If Microsoft wants to push HoloLens to users, it’ll need an app or 50 for that, and it’ll need apps in areas where it rarely pushes the envelope. Microsoft’s core competencies are in Windows and Office, but neither OS nor application are calling out for an AR-enabled headset. If Microsoft wants AR to be the technology of the future, it’ll need to come up with some software that can justify the claim.