Windows, unsurprisingly, gets unstable quickly when I'm running speech recognition, a webcam, email, ICQ, and twenty browser sessions. Hrmph. Someday Linux will have convenient multimedia, and that's the day Microsoft is screwed.

I was talking to a very intelligent friend of mine about how impressed I was by IBM's ViaVoice technology. He had something interesting to say, and it was essentially this: the mouse was not a good match for DOS or for text-based interfaces, since it was an affordance for motion and moving; the graphical user interface was required for a proper "mouse experience," since that is what it is best suited for. Speech, equivalently, is poorly suited to the graphical user interface and is better used with a "natural environment." The keyboard is an artificial interface medium: moving one's fingers up and down in patterns is arguably not a natural way of communicating. The mouse, which replaced it, was a closer metaphor to natural motion (shuffling things around a desk), but was still unnatural. Speech, however, is completely natural, so the jump from an unnatural medium such as the mouse to a natural medium like voice will require an even bigger change in the way we think about human/computer interfaces. The GUI is just not the right thing.

He's a smart lad, smarter than I.

Incidentally, I did a writeup yesterday comparing a bunch of audio codecs: MS Audio v4, MP3, RealAudio, and some others. It's here. Enjoy, and feel free to give feedback or to pass it around.

ACk! I just found out that my mailing list CGI has been broken for months! Why didn't anyone tell me? Damn. Well, it's fixed now, so if you want to add yourself to the list of people to be notified when this site changes, feel free to do so now. =) [slaps head]