I read a fascinating article in last week's Sunday Times Magazine on the subject of robotics, and, in particular, 'sociable robots' -- not just the heavy-duty ones that we're more or less familiar with, doing the repetitive painting and welding on assembly lines and the like, but robots that respond to you; they 'look' at you, 'smile' when you come near, 'droop' when you walk away, and in general exhibit a range of behaviors and characteristics that would lead you to believe that they're alive and responsive - to the point that one group of student researchers, working on one of them, remarked that from one side, they knew that a given robot was a machine, because you could see and manipulate the wires and cables that brought it to 'life', but when you walked around the other side, and saw the big brown eyes and the chest moving as if it were breathing -- well, sometimes it wasn't so obvious.
I've been interested in artificial intelligence for quite some time, as a sidelines observer; I'm not smart enough or experienced enough to understand a lot of it, let alone the intricacies of the sophisticated methods. Even the casual references to AI in this article stumped me, a bit; what the heck is a 'theory of mind', and what point are they trying to make when they talk about it? Are they saying that if they could get a satisfactory one, then these robot truly would 'learn' -- or just that this robot would be able to satisfactorily fake it? But that wasn't a key point, and though I chewed on it for a bit, it really wasn't necessary to be able to understand or appreciate whats being done by some very bright people, both in the direct field of robotics, the indirect one of artificial intelligence, or the allied fields of behavioral psychology and man/machine interaction.
Fascinating stuff. And from a different source, this, which I find simultaneously terrific and creepy:
No comments:
Post a Comment