Artificial life?

Electron · 2000-11-25T08:08:32

Ok, this is not a question that needs a concreet answer. I just want to show my theory and listen to others oppinion about it. At present, many ppl work with developing artificial life. They try to create a program that can learn itself from it''s own mistakes, break the border of logic, and so on. Yet, they only succeded creating a robot (as i know of =) that has the intelligence compared to a child with the age of 3. Of course, you cannot make a program with the same capacity as the human brain - not with present hardware, anyway. For instance, the human brain parts works simultaniously with eachother. the common computer can only work with one part of the time. Thus, i don''t know the number of cells in the human brain, but it''s in the count of bilions. if you wanna write a program that can be compared with the human brain, the program must contain billions of procedures! Good luck, i''ll visit you in about thousand years. However, it''s not impossible! the human brain itself is the evidence of that! So how did the human brain come to this almost-impossible-state? the answer is the only law of nature: Survival of the fittest. let me give you a demonstration. Let''s say you have a creature, that only has one arm and a mouth. He lives in the water. He also have the capability of dividing himself. (like a bacteria). this creature can only reach for the food around him and multiply, until he food drain and he dies. the only way his kind can survive is of multiplying and spread. O -> OO -> OO -> O ->OO ->O ->O -> O O OO OO O ..And so on. notice how the creature''s kind move. the creature itself cannot move, but the kind does. What if one of this creatures born with a ''malfunction''. instead of an arm, he got himself a tail. now he can move around by waving the tail! if the food around him drain, so what? let''s move somewhere else! With this new ability, he will live longer than the others and multiply more times, carrying his new genes to the new ones. Soon the kind will have a new face, and the old will evantually die, or create another speacy. At the next "mutation", a creature will born with both tail and arm. now they both can move for food, and reach the food around him without making energy-draining movements with the tail. now THIS kind is the best kind, and it will push the older away. Now, at the next mutation,the creature had both one arm and one tail, but he had no mouth! this creature didn''t multiply at all. he just died. get the point? the creature with the best capability will push away the less good. and so we have the developing on a new spiecy. Gosh, this is turning out to be a LOONG post, but hang on! i''m at the finish line. My point is: Instead of making billions of procedures, why don''t we just let the computer do it? For example: Make a program that simulates a simple world. Say a 2d world with the dimensions 100x100. One square can contain only one unit at the time. Now, make a food unit, and random some out in the world. Next, make a creature. ABSOLUTELY NO AI!! only the body itself, containing the position of the creature, and a vaule that indicates life. the life value will decrease with one each cycle, but if the creature eats the food, it will gain, say, 3 life. Summon: if the creature doesn''t eat, it will die. Now, make the AI. Make some simple functions, like go left, got up, eat, look down (is there a creature/food on the square velow?) , etc. Now, make the real AI - a matrix containing bytes - that''s all. got you, didn''t i? =) ok, let me explain. say that the matrix look like this: 1, 5, 56, 34, 9, 7, 12, 87, 45, 12 the ai look at the matrix[0] - and recieves ''1''. the ai knows that 1 means "go". go needs parameters, therefore it reads the next byte also. 5 means "left". aha! let''s go left! after he had done this, he reads 56. 56 means "if". need parameters. 34 means "hungry. (life below 10) ok, lets read what i shall do. (he''s not hungry, he will forward the matrix until he finds 12, which means "else".) 9 means "go to matrix number..". parameter indicates ''8'', and so it jumps to matrix[8], and continues reading from there. Now, this example cannot work in practical - it needs MUCH more advanced deconding and some sensing attributes, but the theory is the same. Now, if the creature succeed to eat, and gain some high value in the life attribute, another creaure will born beside him, with a copy of the father''s matrix, but with some randomed adjustements. Now it''s only to lie back and see if some kind of AI is born of theese randomed numbers. *Phew* if you''ve come all the way down here, i congratulate you. Now, i want YOUR oppinion about it! is this a way of facing the mystery of intelligence? Electron"Who need more than 640kb of RAM?" -Bill gates -89

Artificial Intelligence Programming

Started by Electron April 30, 2000 02:52 PM

159 comments, last by Electron 24 years, 2 months ago

Joviex

250

May 04, 2000 04:26 PM

quote:
Original post by NickGA

Just brainstorming, but let''s see.

Computer can: store bits, move bits, alter bits, compare bits (i might be forgetting something here)

Is that all their is to intelligence? Is that all we need? Cool!

No it is not all. What about define groups of bits? You see a car, a toyota. Then you see a bike. What is the difference? They both provide transport, they are both driven? But you can distinguish between both, how will the machine?

When Quality Matters

Send all mail to amorano@bworks.com

gavco98

387

May 04, 2000 04:37 PM

I think we should get some company to donate a server, and run the code on this. Doing this will allow the code to execute 24/7, and we can all view progress by logging in. Running many copys on each individuals computer is relatively inneficient. Im sure companies like IBM will be queuing up to donate a server and be associated with the project.

G Coates
-------------------------------------------
Software is like sex: better when its free!

Gavin Coates
[size="1"]IT Engineer / Web Developer / Aviation Consultant
[size="1"][ Taxiway Alpha ] [ Personal Home Page ]

NickGA

122

May 04, 2000 04:48 PM

quote:
Original post by Joviex

No it is not all. What about define groups of bits? You see a car, a toyota. Then you see a bike. What is the difference? They both provide transport, they are both driven? But you can distinguish between both, how will the machine?

<table width="85" cellspacing="0" cellpadding="0" class="links">
<tr> <td>

When Quality Matters

Send all mail to amorano@bworks.com
</td> </tr>
</table>
</body>

This leads to I believe something called "The Chinese (board?)Room Argument". I read it a long time ago, but if I find it, I will send it here. Anybody hear of something similar.

"If you build it, it will crash."

"If you build it, it will crash."

Joviex

250

May 04, 2000 11:24 PM

quote:
Original post by NickGA
This leads to I believe something called "The Chinese (board?)Room Argument". I read it a long time ago, but if I find it, I will send it here. Anybody hear of something similar.

Never heard of it, but I would like to see it.

I''ll start looking around on the net.

When Quality Matters

Send all mail to amorano@bworks.com

NickGA

122

May 05, 2000 03:25 AM

Here is a link to the Chinese Room Argument by Searle.

http://www.utm.edu/research/iep/c/chineser.htm

Interesting reading if your into AI. It shows the argument and also shows some ideas why the argument may be false. Pretty good Bibilography as well. Check it out

Edited by - NickGA on 5/5/00 3:49:09 AM

"If you build it, it will crash."

MikeD

158

May 05, 2000 06:00 AM

I''ve read most of the URL and I have to say I agree and disagree with Searle. The actions of the man in a room do not constitute intelligence in that format. The man has no idea what he is doing nor what any of the Chinese means so he is just processing information. However, isn''t the basis of almost everything we do, breaking down the real world into symbolic information for processing? Don''t we contain Searle''s man in a box in a lot of our internal reasoning systems. Most of our inputs could be considered pattern matching (vision, sound, touch) with the information parsed into signals our brains can process. But the way our brains process is usually based on pattern matching current situations with old situations and acting exactly the same. We update our neural network with new experiences and then use this knowledge to pattern match better in the next situation.
I don''t think we do anything else.

What I''m almost saying here is that
1)Searle is right, the man in the box is not consciously intentional.
2)We do nothing different on a base level, we just do it a lot more complexly.
3)Therefore, we are nothing more than complex idiots.
(Present company excepted)

Anyone agree,

Mike

126

May 06, 2000 12:25 PM

I agree with you, MikeD. I think that anything in world, does only process information and there is no "you know what you do". We''re nothing more than (complex) machines.

Visit our homepage: www.rarebyte.de.st

GA

Visit our homepage: www.rarebyte.de.stGA

Kylotan

10,513

May 06, 2000 07:40 PM

How can you say there is no "you know what you do"? When I do something, I can think about it and understand it. Whatever you want to call that, it is a level above merely performing the function itself...

Joviex

250

May 06, 2000 10:55 PM

quote:
Original post by Kylotan

How can you say there is no "you know what you do"? When I do something, I can think about it and understand it. Whatever you want to call that, it is a level above merely performing the function itself...

this is the very defintion of self-aweness. As far as I know, the defintion of true life is intelligence, self-awareness and consciousness. Now, the first two are easy to figure out, it is the third that has man baffled for 50,000 years now.

I also agree. I read the article and I don''t think it really proves anything either way.

Man relies on symbols for storage (sights, sounds, concepts). The very nature of this makes it next to impossible for coimputers of the silicon variety to exhibit "life". They can draw a close bead to its beaviors (i.e. with logical rules) but when faced with simple deduction (i.e. pick a woman''s face out of a crowd of men for example) it will fail horribly, simply because a face is a face is a face to a machine. How can it know the essence of the face?

This "search for intelligence" in other forms is indicative more for the search of one''s soul and meaning on this dusty sphere.

So, I do agree you can make a nice static environment on your truely binary system and emulate much of lifes behavior patterns, but the machine will never turn to you one day and say it is bored.

When Quality Matters

Send all mail to amorano@bworks.com

126

May 07, 2000 03:40 AM

quote:
Original post by Kylotan

How can you say there is no "you know what you do"? When I do something, I can think about it and understand it. Whatever you want to call that, it is a level above merely performing the function itself...

Probably most people think that they think on their own and make free decisions. But what we call thinking or understanding is nothing else than some physical reactions in our brains.

Visit our homepage: www.rarebyte.de.st

GA

Visit our homepage: www.rarebyte.de.stGA

Artificial life?

This topic is closed to new replies.

Popular Topics

Recommended Tutorials

Artificial life?

This topic is closed to new replies.

Popular Topics

Recommended Tutorials

Reticulating splines