Perhaps this is more of a philisophical question. But what, exactly, is "Artificial Intelligence " ? What makes it different from our conception of "Intelligence " ? Does the word "Artificial " have any significance whatsoever, or is it merely a figment of our imagination ? Are humans intelligent ? Or is it merely an illusion ? For those who believe in creationism in one of its various forms: if humans were created by a/the "God" entity/force/law of the universe, are humans not "Artificially Intelligent" ? If we created AI, would using it be ethical ? Or would it be just another form of slavery -- the intentional limitations of the program the bondages of servitude ? Would AI be a new species ? If we didn''t get along with this new species, what then ? Even if we did get along, would we be comfortable treating AI as equals ? Or perhaps AI would even be superior, leaving humans at their mercy ? Maybe the AI doesn''t like working either, so it has humans do all the hard labor ? ---------------------------------------------- The last paragraph prolly consists of well-addressed issues, but there are so many questions. Anyone else find it odd that we are trying to recreate ourselves, even when we don''t really know what we are "That''''s to do with the day I finally realized the world had gone totally mad and built the Asylum to put it in, poor thing, and hoped it would get better." -Wonko the Sane So Long, and Thanks for All the Fish

Can AI go insane ?

Is it possible for any sentient being to decend to such a low level in its "brain" that it can do simple arithmetic as efficiently as a non-sentient pocket calculator ?

I''m afraid to post many questions -- I feal like I need a separate forum for the "As-of-yet-uneducated" to babble about whatever comes to mind, and not worry that someone else has already answered that question at some point in time, somewhere in the world.

I should just get NOOB tattooed to my forhead and be done with it.

You''re still reading. Your tolerance for my insignificance is noteworthy. I wouldn''t be overly offended if someone deleted this thread.

it''s only "artificial" because anything mankind creates is termed "artificial" in the english language. It would still be a new life-form if it were to become "intelligent". Though in the end we will all just be consciousnesses in a giant computer because it would eliminate a lot of problems, mostly mortality. ("Diaspora" by Greg Egan). And if we did create an intelligence that exists in solid state, there would be no reason to assume that it would take on human characteristics such as laziness and insanity. Though read the short story "I have no mouth and I must scream" by Harlan Ellison, as this offers a view on why AI should not be created.

You can raise main filosofical questions about intelligence, but to define "Artificial Intelligence", you don''t need to.

There is a very simple definition (that matches perfectly what AI researchers are doing): "Artificial Intelligence is the subfield of computer science that tries to make the computers do what we still don''t know how to make them to do"

If you want a definition of the work "Intelligence", I can give you one that I specially like (It''s not exactly a definition, but expresses clearly what are the problems to define Intelligence): Humans consider intelligent to those reasonings procedures that they cannot understand. Once a reasoning method is completely understood, and can me computerized, it''s not considered intelligent. Think of: automatic theorem proving, playing chess, etc. all those problems were considered as "needing intelligence" problems, but today you can find algorithms to solve those problems, so we consider that there is no need of intelligence to solve them. Therefore, my conclusion is: "We think that we are very special for being intelligent beings. Therefore, we will not consider intelligent any task that a computer can do. But in fact, no one can give a good definition of what intelligence is".

I''m sure that if one could peek into a human brain and read what is the definition of intelligence stored there, it will read something like: ''intelligence is what a human brain has'' and nothing else. The probles is exactly the same as defining the word "life".

quote:

---

Humans consider intelligent to those reasonings procedures that they cannot understand.

---

Very good definition actually. It takes alot of intelligence to develop calculus, but once it is done, repeating it isn''t intelligent. Applying the proper setup to solve a problem is intelligent. Using a "precedent" or previously used setup to a problem isn''t intelligence. There is a huge difference between being intelligence and being a mimic. Which makes ya wonder how many people in the word qualify as being intelligent. =)

And to answer the question, can AI go insane, of course it does. The first time I implemented my tank AI, he drove in circles and shot the ground, killing himself. Pretty crazy if ya ask me. Even now, he will drive around and chase and shoot, but every now and then drive off and shott wildly into the air. "Insane" is a weird word. Even in human terms, there is a fine line between an insane person and a genius. If my AI tank decided to stop chasing and go in complete different direction, I would think it is broke/insane. But when he ends up just taking a shortcut and planting mines in the target''s path, he is genius.

So yes, an AI can go insane. However, if you program it to appear insane, that is not the same thing.