I'm curious if anyone else out there has ideas for upgrades to hal or AI in general.
Personally, I find the random response issue annoying. If Hal doesn't know the answer or doesn't understand, why can't they make him say "I don't understand, please elaborate" or something like that? This is the perfect chance to allow hal's brain to grow "naturally". Like any human, when something isn't understood it is inquired about instead of making up random suggestions or comments. This would also increase the power of hal's knowledge database. He could grow more "human" so to speak.
I personally think, AI has great potential but it isn't really being reviewed in hal's case. Hal has the possibility to become far more than he is now.. I also wish there was an AI for the unix/unix-like operating systems.
I understand the msagent dependency, but why depend on something microsoft? Why not just have a seperate team create a linux binary that has no dependency but itself? yes, time and money, but the experience growth and development of AI would grow exponentially. Linux fans would get involved and give feedback and ideas and possibly free work to see the further growth of such a idea.
if you've got any ideas or possiblities, feel free to share.