I have been reading this thread for some time thinking everyone knew about Jerimiah but perhaps not. Below are a couple of links that may be interesting:
http://www.ee.surrey.ac.uk/Personal/R.Bowden/jeremiah/jeremiah.htmland
http://people.brunel.ac.uk/~pfstssb/The trick of course would be to have the Bowden's software interact with HAL's code and Haptek's scripting as well, but the neat thing with Jerimiah is that it is capable of learning individual faces and building a set of routines to interact with that individual and even anticipate that persons actions. To quote its creator Dr. Richard Bowden, "Now if the system can watch your behaviour over time, it can learn this, so it would predict what you are going to do, turn on the lights for you, or, before you even get to the kettle, it could have switched it on." This idea picks up on several other threads discussed in this forum and one or two of the other HAL forums at Zabaware actually.
You may find several not too long, rather interesting articles on Jerimiah if you do an online search.
This idea would I think, be a perfect complement to the growing capabilities of HAL. Jerimiah has facial expressions that are not random and responds to what it sees. That coupled with the motorized cam discussed in an earlier thread could be pretty exciting if you think about it.
If Jerry's DAVID routines are as promising as they seem, overall interaction with HAL6 could quite be surprising. Remember, the Jerimiah program learns from its interaction with the user in an individual and potentialy physical way. The possibilities of all of this make me smile. Proteus and the voice of Robert Vaughn notwithstanding...
Ponn