I assume there's a good amount of neuroscience involved, too. I mean we're still not entirely sure how we learn, meaning exactly what changes happen in the brain. How do you build a synthetic system that learns if you don't understand how it takes place? Or is there some computer science work around for this?
The interesting thing we've been able to glean about artificial intelligence is that it is much more effectively allied to being a specific intelligence, rather than a general one. We can teach a machine to do a small subset of tasks very effectively (and can even improve itself) with a high level of "intelligence," but we've had pretty dismal results in making one that can behave intelligently in a wide variety of situations, even to the level a toddler interacts with the world.
Very fascinating subject.
Hey Jimmy, you gonna throw up some "Let's Talk About" threads this time around in the Andromeda forum? I mean I know it's early, but I like to know I have something to look forward to.
Hmmmm. The thought hadn't even crossed my mind, actually.
A snarky part of me is thinking "but I don't really WANT to talk about ME:A." I'm going to follow the development at a high level like I do lots of games, but I honestly am not likely to play it unless something TRULY interesting presents itself. I'm kind of shut off on the entire franchise, so I can't really see myself trying to begin a game development dialogue in those forums... simply because I have absolutely dismal view on the game itself. Which may not be fair, but that is my mindset at this time.