Saint Satan, I'm sure that the model you suggest is going to be tried but so will the goal of making a conscious machine. Humans are just that way, they are going to try all the possibilites. Actually your post reminded me a bit of a famous science fiction writer (Asimov?) who came up with the "laws of robotics" so that they would never hurt a human.
D wiltshire, right now A.I. is being used in many applications, including corporate profits. It will be interesting to see what happens when our mind children also have free will and pain.
2SYN, I don't know how much you have been keeping up with A.I. but it is a whole different ball game these days and the paradigm has changed dramatically in the way to go about making intelligent machines. There is NO programing now. Neural nets learn to see for themselves by interaction with the real world, computers learn to walk by themselves (see Honda's incredible new walking robots for an example of this, they can even walk up and down stairs and look remarkably human doing so). Thier intelligence is still small in terms of direct comparisons with biological systems but are growing rapidly. They will reach human levels in a much shorter time than people expect.
The big thing though that divides the A.I. community is if they will be able to gain "conciousness." Some people, perhaps such as yourself, believe that more is needed than the hardware or wetware. I've seen arguements on both sides and I don't think the jury is in yet. Even so, computers are heading in the direction of the quantum world. They have already done quantum computing and in another 10-30 years we may see a "chip" (probably won't look anything like conventional chips used in todays computers) that will be smaller than a pin head that will have more computational power than the human brain. Both fascinating and scary if you ask me.
Cello, if there are typo's it is probably my own fault. I didn't run it through a spell checker. There are actually some cool programs I've used for cellular automata or sometimes called "artificial life." Some of those evolutionary algorhtims are pretty amazing as well, if nothing else they are quite entertaining.
Scully, I actually wondered if a robot can cook and yes the Japanese have developed a few models that they have in commercial use in both homes for handicapped people and in some cafes. Before long, McDonald's won't be hiring j-dubs to cook their food any longer and maybe then I can get my Big Mac the way I like it. haha
Hi Hmmm, yes just brute computational power doesn't mean intelligence. It is only the engine for the software of course. The software in this case is not the old programing style of expert systems in LISP or other AI languages but is self evolved. Will they be able to get to where they can pass the Turing test? Time will tell. If these stories are any indications though, it may be an interesting test indeed.
Kind Regards,
Skipper
