Ive been looking over some of the older posts in here, some talk about "Today's Technology" etc etc. I know I am going to get yelled at because of this, but in respect to VR, Charles, I dont think you are up to date on todays technology. Have you ever used an application that utilized MS SAPI 5.0? I know VoodooVoice used 4.0, which I will admit took some processing power and wasnt very accurate. I won't even go into lower versions. However, SAPI 5 is MUCH better than SAPI 4, it is faster (better loading times as well as recognition times), and has higher accuracy. It also has more features. It doesnt need any special hardware (DSP cards etc), or even all that fast of a processor (as Mellman put it earlier, if you can't afford a K6-2 450, why are you building a carputer?). Ok so say you don't like MS (me included ). There are plenty of alternatives. I havent looked around, but I think IBM has something, plus there is ViaVoice (if i got stuff wrong there dont yell at me, I know there are alternatives, I just havent looked into them).
So as far as technology goes, voice recognition has come a LONG way. The algorithms behind it are getting more accurate and faster.
You can contact me on AIM @ DustDvlKC if you want to discuss the algorithms and such more - I have done a little work with them, and while I'm no expert, I have a basic understanding of how they work.
I hope smashr made the video today, because otherwise this argument is going nowhere really fast.