Some interesting and cool demonstrations of what mouser pointed it is largely "old" technologies, but converged (supposedly) into a more cohesive and functional device. The thing is I don't see how many of the things demonstrated in the videos actually work, and until I see 3rd party demonstration of it (e.g. by an independent reviewer), I'm rather skeptical.
For example the "copying" of things from the real world on a piece of paper into your projection system, sure I can see that, just take an image and overlay it in your display. There is one problem with this, which is how do you separate background from the object you want to copy - in the example shown (a pie graph) it's a white background, so maybe that's a requirement. That would be rather limiting if so, but better than nothing. I can believe it working in the case shown.
What I don't trust as much is when he then takes that object and adjust the ratios of the graph! That is a much more difficult thing to accomplish, and I have my doubts it was done in any sort of dynamic way, rather than the more likely mocked-up or specially coded demo. Even if it is really working that way, it is surely with a very small set of use cases, and yet it is presented with the idea that it is showing this incredibly flexible tool; with the implication that you could interact with almost anything this way. In reality
In general the whole thing is a bit curious too, because he shows the helmet with the projector and camera early on, and then later all you see is a small thing around his neck. OK, there are micro projectors now, the size is not a problem. But what about all the incredibly odd angles that things would need to be projected on throughout the video? You always see the projected image, or him, never really both at once (except parts obviously, like his hand or arm). It's never really showing how the projector is positioned *as* it projects, and how it is aligning with the surface. Certainly there are corrective measures that can be taken on image output, but to believe they can be that extreme (the device around his neck was not fixed in place, so it could be at all sorts of odd angles), and done that quickly, is a bit of a stretch to my mind.
Interesting, but waiting for further proof of capabilities...Edit:
Re-watching the video, for the most part it looks like he puts objects down in range of where the projector *could* project, but the angles are still often odd. The use cases are limiting for this reason alone. And I imagine it might end up affecting your posture if the only way you could change the position of your display was to shift your body around.
There is also another TED "talk" where his teacher or some other MIT person also "demos" the technology... but what's odd is that she claims to be wearing the device around her neck which, presumably, works (and costs only $350), yet you never see her turn it on or use it or actually *demonstrate* it. Everything in the "demo" is just the same clips from the original demo. I guess maybe this is TED US vs. TEDIndia, and that's why it is "presented" again, but it does seem odd that two presentations were made, neither one of which with any practical demonstration of the technology.