"The main thing probably would be the resolution," he said. "Currently your hand is a point in Kinect. We'd like to see the whole hand being broken down into each joint so that it would detect your hand movements.
"Also, increase the AI of the voice recognition as well, because it's relatively primitive at the moment.
"We could do with some improved depth, so X, Y and Z. Lastly, possibly, the field of view. At the moment you have to be quite far away from the screen because it doesn't… you need two metres, I think, at least - if they made that shorter and maybe detected a wider field of view, then I think it would be nicer for most households. Especially in Japan where you don't have five-metre living rooms."