Hey Matt, I'm on "tour" right now (no email responses 'till next week btw for people who have sent mail to qflux), but some quick responses from somewhere in Montana.What I'm coding is this...I have successfully integrated the Microsoft SAPI5 with the VIM, which is their new speech synthesis and recognition system included with Windows XP and installable on other versions. I'm feeling these days that language is going to be the key method for the VIM to express its intellect. So the first stage will be a stand-alone program in which the VIM will read to you...VIReader. The graphics is a very realistic version of the female characters face, which will morph to the sounds spoken (text-to-speech) of a file you load. The role of the matrix will be to set the expression to the text; the eyebrows, mouth and head movements responding to the content of the text. That is about 50% done. Then, the goal will be to have the VIM, after extensive training of a large matrix (done by reading many books and dialogs), respond on its own to sentences you type, speaking back it's own responses to your dialog. Then using the ms voice recognition enable the system for two-way verbal communication. Finally, by linking matrixes that different people train, we'll have a full-fledged machine intelligence. So like Max Headroom crossed with HAL.