> I don't understand the insistence on using voice as the main interaction and ditching the screen.
There was a google i/o talk a few years back were they talked about users wanting multi-modal, an example being they ask for restaurant recommendations by voice, then get the list they can view on their device. Both query and results are presented in their easiest modal, and humans will naturally switch between them.
This thing seem dead on arrival. Who wants to hold their hand up like that? Who wants to look at an uneven "screen"? Can you use it while walking or experience the movement in a vehicle? (car, bus, subway)
I agree with you on all points except one. Arguably the uneven "screen" problem can be solved with a depth camera and warping the projection to match the contours of your hand. Since they already support hand gestures on the target hand it's possible they already have the equipment built-in to do this.
If you've used VR passthrough, you'll know that the latency will not work out well. You have to capture, analyze, calculate, and project in low, single digit milliseconds, against a moving target from a moving target
Eyes should stay mostly fixed relative to a pin on your chest, as long as you're looking in the same direction (ie your hand). I think the differences would be small enough, especially since the display seems pretty blurry anyway
Projection warping should be viewer position independent (within a limited scope), only relying on the position of the projector and the screen. Some home theater projectors which already do this to an extent. And there are some famous public performances where video is projected onto a building and warped to match the contours in such a way as to give a convincing 3d effect (Example: https://www.youtube.com/watch?v=gJ_5sDvAlNY). You can perform this kind of projection warping on a raspberry pi.
There was a google i/o talk a few years back were they talked about users wanting multi-modal, an example being they ask for restaurant recommendations by voice, then get the list they can view on their device. Both query and results are presented in their easiest modal, and humans will naturally switch between them.
This thing seem dead on arrival. Who wants to hold their hand up like that? Who wants to look at an uneven "screen"? Can you use it while walking or experience the movement in a vehicle? (car, bus, subway)
Is this just a big sunk cost fallacy launch?