The Speculist

Nah, it just means they'll develop their own senses of esthetics, or perhaps that they won't have any need for them. They can still safely build Terminators and go after us. ;)

Seriously though, I think you worry too much about machines taking over. I think it is anthropomorphizing them to assume they'll have any ambition to do so. That's one thing that would be hard to engineer into them and entirely unuseful for them to have.

-Jim

Jim:

Ahnuld Robot voice:

"Truth is beauty, beauty truth. That is all you know and all you need to know...John Connor"

Jim,

Well, I don't really worry about it. Much.

It's interesting that you suggest that I'm anthropomorphizing computers. If, in fact, AI is achieved by reverse-engineering the human brain, then the new machines might have a lot of ideas and ambitions that we don't think of as optimal for equipment that we keep around the house, but that are part and parcel of being a thinking being.

As a science fiction writer (alas, unpublished as yet) I have to say that there is a temptation to view that anything which is possible (artificial intelligence, flying cars) will be done in large enough quantity to change the world. The truth is, inventions that change the world have to be possible, *useful*, and in modern economies, manufacturable and salable. Remember the earlier discussions of flying cars where it was concluded, more or less, that while it is possible to build prototypes today, manufacture, sale, and mass use are utterly impractical now and at least a certain distance into the future.

As a result of these constraints, other than in academic pursuits, I don't think we're likely to see generalized human level artificial intelligence. Not because we won't ever be able to produce them, but because they won't have any particular market value, so they won't exist in large numbers.

A servant robot really need only have dog intelligence. Understand commands. Do what you're told. A dog with hands and speech would be a very useful (but controllable) level of intelligence indeed. Likewise, an intelligent car need only have rat intelligence to do all the 'intelligent' functions that are useful - come when called, go where told, not run into things and dodge things running into it. Heck, cockroaches do this.

An artificial intelligence can be, and I think will be, the ultimate specialist. 21B, for example, the medical robot from Empire Strikes Back, need not know about or care about anything in the galaxy beyond his surgery. (His, in this case, because he had a masculine voice). He need not have emotions beyond empathy for his patients, he need not have fear, he need not have a sex drive (heaven forfend). His will be very much an alien mindset, but he is *product*. He fills a market need. Nobody would want a surgical robot with a hobby.

I think this is the direction artificial intelligence will take. We humans are best at being generalists. The brains we build will be specialists.

Of course, there is the risk that we will build specialist killer robots. Terminators, again, leap to mind. When we do so, we must be as cautious with how we deploy them as we are with other potentially species-destructive weapon systems, or we must realize that again, a soldier robot probably doesn't need much more than dog intelligence.

By the by, I do not by any means suggest that current soldiers are stupid, merely that the brainpower of a person with a rifle is rather badly under-utilized, and may even get in the way.

-Jim

Evaluate beauty... I should really stop carping and just be glad that we're still good for something.

At least in certain domains of art. There's been some research with abstract art, such as Jackson Pollock's drip paintings, which indicates that artwork with certain fractal dimensions is visually appealing.

Fractal Expressionism: Can Science Be Used To Further Our Understanding Of Art?

Similar research was done studying Gestalt principles and fractals in Japanese Zen gardens:

Visual Perception of Japanese Gardens