[User9812]

I thought that was self-explanatory. Your iPad camera faces forward, the mirror redirects it down to the playing surface. The AI tries to track the location of the letters and colored blocks in the mirrored image.

What does it do? The examples are pretty obvious, what more are you looking for?

Arrange tangible puzzle pieces into matching on‑screen shapes.

Be the first to guess and spell out the on‑screen hidden word by tossing down real‑life letters faster than your friends. A related picture gives the clue.

[mikeash]

Maybe the problem is just that what it does is so far removed from the superlative language used to describe the product as a whole? The introduction leaves me with the impression that it must be more than a silly augmented reality puzzle game.

I still don't understand the mirror. iPads have a camera in the back already....

[User9812]

http://i.imgur.com/0OXmc9h.png

The blue line is an iPad, the red region is the camera line of sight. On the left, you have no mirror, the camera is looking straight ahead at you. On the right, a mirror is attached, it redirects the camera view down to the playing surface.

[mikeash]

Got it, thanks. Nice diagram. I can't help but think that if the connection between the iPad and the stuff in front of it was made a bit more clear on the site, as you have done, that it would be much easier to understand. Or maybe it's just me.