While I agree about voice (and that even though I've been working on TTS for 6 years now ;)), I don't think screens and keyboards should be the ultimate goal, but rather something like brain-computer interfaces.