I have entirely different ideas about how I want to radically change the way that my
interaction with my computer happens.
Our problem is that it's an enormous amount of work to even prototype novel interaction styles
(or whatever you want to call them). It shouldn't be. I'm not sure where we went wrong
though.