I played with Hololens at Build this week. I believe, after seeing it, using it, and even developing a little with it, that this device is truly revolutionary, and Google is wasting its money in Florida.
I shook my head back and forth like a dog trying to dry itself off, and the images I saw barely wiggled. The images are bright enough to completely occlude reality. The software dev stuff is from MS, so it is polished, simple, and powerful.
Voice recognition on the device, for example, is just handled by calling the voice functions in the core library, then GIVING IT A STRING. It listens for this plain text string. That's it. Sooooo simple and powerful.
First-mover is an advantage, sure, but it hardly means that anyone who isn't first to the gate is wasting their money.
Anyway, I see downthread that VonGuard has made the silly claim that "Hololens can do everything Magic Leap is supposed to do", so I think it's safe to say that the "wasting money" thing is just a throwaway line that's not based on a lot of actual knowledge in the field. I feel kind of trolled.
I think what's happened is that you meant "supposed to do" as in "it fills the need that ML was aiming at", and I took as "it has feature parity". The latter is clearly not true - ML is supposed to have a very wide FOV, handle the accomodation reflex, simulate occlusion, do depth of field, etc. - none of which Hololens can do.
Alexa Skills kit voice recognition configuration is also done by just giving it a String. It's a wonderful time to be working on apps/APIs using voice recognition!
Magic Leap is based on super complex tech that draws images right onto your retina. It's the kind of tech that will take them another couple of years to make small enough to package, and possibly another few years to build the tools to develop for it and to ensure it's not actually blinding people by shining bright lights right into their eye. A friend of mine used it and said Magic Leap, initially, is like trying to breath liquid oxygen in the Abyss: you fight it and can't really take that bright ass light shining in your eye. You have to really get used to it and relax yourself for it to work.
Hololens is just a screen in front of your face. It could never blind you or hurt your eyes. It just works, it works today, it fits in a product form factor today, and after using it for a while, I really don't see much need for an also-ran here. Hololens can do everything Magic Leap is supposed to do, and it can do it a LOT cheaper and safer. Plus, MS dev tools. You just know that Google's Magic Leap dev kit will, basically, be targeted at super intelligent people and require a lot of do it yourself stuff.
I havnt used either, but my impression was that Magic Leap's technology is basically projecting a light field into your eyes, so you eyes can actually focus naturally.. as opposed to looking at an image with fixed focal distance like the hololens.
But I'm not super informed about either product so I could be wrong.
Thanks for the reply, even though I was hoping it wasn't basically hearsay and speculation (which it unfortunately is). Abovitz has repeatedly said that they absolutely want their tech to be safe enough for kids to use, so perhaps the safety concerns are not based on facts? Magic Leap apparently has active occlusion, and can run the spectrum from VR to AR/MR; pretty sure that makes it the potential leader, not an also-ran. I guess we'll see soon enough. Exciting times.
What I'm saying is, Hololens does ALL of those VR/AR things already, and there's not even a slight chance of eye damage involved. I really have trouble envisioning anything that would be lightyears beyond Hololens, but rather, something that would be a small step up from Hololens, and yet, still 2 years out.
Jut wrong about Hololens. Stuff may be drawn close to your face, but the image is scaled very, very well. It may be drawn up close, but it still looks like it's very far away, way more than 2 meters.
I remember it being mentioned in Wired in the 90's. But then, Westinghouse build a Mech (yes, a battle suit) in the 50's, and they never caught on. Of course, they never put anyone inside the mech because it could have snapped an arm off by bending the wrong way...
Some tech comes out early but doesn't catch on for a reason.
I shook my head back and forth like a dog trying to dry itself off, and the images I saw barely wiggled. The images are bright enough to completely occlude reality. The software dev stuff is from MS, so it is polished, simple, and powerful.
Voice recognition on the device, for example, is just handled by calling the voice functions in the core library, then GIVING IT A STRING. It listens for this plain text string. That's it. Sooooo simple and powerful.