Creating a Digital Consciousness

Most of those things you mentioned are only appearance differences, not functional differences. A computer could theoretically be programmed to do any of those things.

 

.

 

To be getting into the philosophy of it all: Perhaps all things have a form of consciousness, we as humans, are unaware of due to our genetic, our genetic disposition.

At any rate, and after much talk between us all here, we can hopefully agree it -is- potentially possible to create a consciousness.

Digital Conciousness would require, I believe, the will to grow.. Or the natural ability to grow. Consciousness is dynamic and creative (to certain degrees). For a synthetic brain to have this would be a sure-fire sign we're heading in the general direction of consciousness.

 

Oops, double post

 

.

 

.

 

I don't want to sound like a smart ass, but we already create new brains out of biologic material (aka babies). We even program those brains (aka raising, training, teaching). Those brains can program themselves.

I thought the challange was to recreate (or at least mimic) some kind of consiousness using non-biological materials.

 

Fair enough mate. If you want to beleive that your brain is just a big deterministic computer, that's fine by me! ;-P

 

I never said 'deterministic'. Teaching and learning, with its imperfections and indeterminism, is the way that mammal's brains are "programmed."

Determinism is a feature of our current methods of "teaching" computers. There are two thrusts in AI.

One is to build a non-deterministic machine.

The other thrust is towards programming deterministic machines in such a complex manner that we don't know whether the result is deterministic or not. Self-programing, heuristic (learning from experience) programs can get fiercely complex.

 

The comparison of a computer to a brain, or vice versa, is just that...a comparison. I think that what's important is the functionality, not the appearance. A computer could theoretically be LIKE a brain in every functional way and still not be a literal brain, as it isn't a pinkish organ in some animal's head.

None of the aspects you described are fundamental differences between computers and brains that can't be overcome with the proper technology. It's true that one can't easily isolate memory in a brain as one can in a computer...but that's just a limit of our current neurological technology, not a fundamental difference. Same with self-repair...a limit of our present-day computer technology, but not something that computers could never theoretically be built to do. Some of the differences (computers operating entirely by electricity, brain using chemicals AND electricity) are merely superficial differences in the "engineering" of the two. Yes, a computer operates solely by on/off switches and a brain is slightly more (but not much more) complex than that...but that just means it might take more than one on/off switch to simulate a neuron's behavior.

 

So how do you make a non-deterministic machine then? With programmed randomness? D'oh!

 

Any machine, such as a computer or a brain, will SEEM non-deterministic if it is complex enough. It's true that the actions of even a computer as complex as a brain could theoretically be predicted if you knew every variable...but the same is true of a human brain.

 

analog would be easier, they had damn good ai in the 70's w\ analog computers, analog is cool, you could do it w\ lasers and glass cubes

 

.

 

.

 

.

 

(quote)
could theoretically be predicted if you knew every variable...but the same is true of a human brain.
(end quote)

Think hard about whether you really believe that...

 

The problem is that we DON'T know every variable. Theoretically, if you had a perfect grasp of neurology, a full map of someone's brain, and an unimaginably fast computational ability, there's no reason you wouldn't be able to predict the response to any stimulus in the short-term. We can already (crudely) predict how single-celled organisms react to stimuli. The only difference is the complexity.

To accurately predict the actions of a human based on their brain map would require an intelligence MUCH more complex than the human brain, just as predicting the actions of single-celled organisms requires an intelligence MUCH more complex than theirs. Any complex system will seem deterministic to an entity of much greater complexity, and any complex system will seem non-deterministic to an entity of lesser, equal, or slightly greater complexity. That's why we can't predict the actions of other humans...they are just as complex as us.

 

Alan turing, got screwed up the ass because he allegedly was involved in screwing up the ass, the fucker saved the world and his gov't killed him happy shit huh?

 

i believe that artificial intelligence will only come around once the human brain has been understood.

i dont think its a matter of the structure of the brain, moreso, the functions that the brain does. while there are increidbly diverse networks within the brain, there is a not so increidbly diverse number of chemicals and cell types that make up this network.

the network itself is not grown in one go. the networks must be developed over time, with learning. conscious thought is only possible through learning. while a baby is in the womb, it learns its basic instincts, and reactions. then the baby is born, and it is as intelligent as any other new born animal. however most animals learn practically their WHOLE mind while in the womb and so adult animals arent much more intelligent than younger animals (except in things that require experience of course, like recognising things).

artificial intelligence will therefor require the mimicing ofd the individual systems of the brain (there are around 200 differnet types of cells int ehw hole body, so it shouldnt be impossible to recreate the main functions of the brain).

the hardest thing to recreate in terms of intelligence, i think, is going to be the ability to recognise. recognise objects from diferent angles, percieve surroundings.

the thing is, as i mentioned in anotehr thread is that i dont believe we can achieve artificial intelligence before understanding our own brains. science has uncovered a very small understanding of how the brain actually works, most studies are on how the brain thinks, and not what is actually going on with each thought.

and i think the onl way that will happen is through studying the effects and pharmology of drugs.

so i think drugs will lead to artificial intelligence, and through no other pathway will an organic mind be replicated.


a real AI computer will have very very few things programmed into it, and the process of creating the intelligence will be the teaching aspect.

for example, the only accurate speech-understanding computer will have to be tought a language from scratch in order to develop and percieve language like humans.