artificial intelligence

[Cythraul]

Sorry if there's already been a thread about this but I can't be bothered to search for it. That notwithstanding, I'll pose a couple very general questions: Do you think that it's possible, in principle, for a computational device to be conscious in the way that humans are? That is to say, do you think it's possible for there to be "something it is like" to be a computer? If your answer is no, then in what way is human consciousness fundamentally different? I'm not very up-to-date on the most current research but I'm familiar with most of the basic concepts and arguments in the literature. Also, a further question: if a machine was conscious how could we tell, given the difficulty (perhaps impossibility) of observing first-person conscious experience?

 

[Demiurge]

If your answer is no, then in what way is human consciousness fundamentally different?

The ambiguity of its determination.

 

[Kenneth R.]

human decisions are not always based on logic.

 

[Voice of God]

In the future, yes, I believe that a machine can be conscious (that is: aware of it's own existence and "thinking") and very much humanlike it other ways too. Neural networks and similar techniques exist already and I think something very much like human conciousness can be taught and programmed into a machine. We're just not there yet tech-wise. Most likely it won't be like human on every aspect, like SS said decisions are not always based on logic (although it raises the question, can emotions be programmed? I don't see why not.), but as for consciousness, it has only little to do with decision-making.

 

[Hypnos]

the big problem is, we don't know what consciousness really is and how it works. it is still undiscovered what the origin of consciousness is and which neural process is creating it...

 

[Silver Incubus]

yeah it can only really be simulated consciousness at best.

 

[speed]

I know Roger Penrose in An Emperor's New Mind, claimed conscious AI was impossible. However, technology has come along way since then. What are the new views? I really dont know. I am a luddite who would prefer research into AI to stop before it is too late, and computers act like any other living orgnanism. And lets face it, the animal world is one of mass murder.

 

[Kenneth R.]

i would not disbelieve claims of capability to create thinking, adapting computer AI. however, the major difference is that a computer remembers everything it's learned (unless it voluntarily deletes information, which it can then never recall again). and that no matter how many emotions are programmed and how effectively, the machine will still react to them predictably because it will use the emotional programming to make decisions. it won't ever make a "wrong" or irrational decision based on inputs it was programmed to react to.

 

[Iridium]

So is the essence of consciousness unpredictability? The only reason it's possible and even relatively easy to predict what a machine will do is because we have easy access to its memory and thought processes. Since we've created the intelligence, we can manipulate it. What we designate as emotions are merely functions of the reptillian brain, a structure that evolved far before our species. There is a fully functional robot at my mother's workplace, which formulates ideas and speaks at the level of an eight-year old, and was able to recognize itself in a photo. Is that not artificial intelligence? Perhaps our definition of consciousness is clouded and incomplete, which is why this question even comes up.

 

[Voice of God]

and that no matter how many emotions are programmed and how effectively, the machine will still react to them predictably because it will use the emotional programming to make decisions.

How does this differ from humans? People also use their emotional programming. If we could know ALL the "programs" and processes of how a certain person deals with things and ALL the environmental variables, his/her actions would be predictable too. With computer intelligence, we usually do know these things.

But like Iridium said, the programming on a robot is easily accessible. The difference between human and (current) AI processes are that most human processes are unconsciously learnt from the environment the person grows up in and they constantly evolve and the AI is pre-programmed.
I believe it is possible to create an AI which does the same thing, creates it's own processes (cognitive, emotional and others) when needed and evolves them with the environment possibly diverting from it's original programming if necessary. And with time the programming of such AI would be based on the original design, but most likely as complex and unpredictable as the human mind, except that it would be described in a language that we can understand and eventually "solve". Such is the case with brains too, we just don't have good enough grasp of the brains' "language".
(btw. I don't think such an AI would necessarilly be a good thing to create, but I believe it could be done eventually.)

And we do predict peoples decisions and reactions, all the time. Not with 100% accuracy, because we only know part of their process and because our own processes are not necessarily optimized for that purpose. Everytime you pick a present your girlfriend likes, you've predicted her reaction. Everytime marketing people come up with a product that everybody must have, they have predicted (and usually manipulated) people's decisions.

 

[Birkenau]

i would not disbelieve claims of capability to create thinking, adapting computer AI. however, the major difference is that a computer remembers everything it's learned (unless it voluntarily deletes information, which it can then never recall again). and that no matter how many emotions are programmed and how effectively, the machine will still react to them predictably because it will use the emotional programming to make decisions. it won't ever make a "wrong" or irrational decision based on inputs it was programmed to react to.

Don't people tend to do what they've been 'programmed' (by programmed, I mean the enviroment they've been raised in and the people they've met) to do? A person is only going to make a decision from past experiences and gained logic, which is really the same as a programmer who updates an AI software to fix bugs, faults etc. from past problems and exploits.

I don't think there is really much difference between a robot and a man now days, most people are just doing what they've been programmed to do from major sources (mostly television unfortunately), and have no consicious of their own, of course, there's always a few dissents

 

[Hypnos]

the main difference between humans (and all organic lifeforms) and computers is: we work with analogue signals while computers are based on binary data. this could be a big step to take in creating AI.

 

[Cythraul]

Here's a link to a paper by John Searle. I tend to agree with him on the issue of AI. http://www.ecs.soton.ac.uk/~harnad/Papers/Py104/searle.comp.html

 

[Dushan S]

I think that whole question of Artificial Intelligence is actually asked in a wrong way.
Starting point is usually idea that we are our intellect and that what we need to do is recreate intellectual mechanism of human being. Then, someone can point out that human is not human if he has not emotions, and then maybe we could make some chemical system that would imitate chemical reactions in our body. We could even make some kind of body that will have all 5 senses and give information’s to the brain. But it seems that more or less intentionally scientist are avoiding to speak about one fact, partly because science is not able to comprehend it, and it is still in domain of spirituality: awareness

It is quite normal and usual, and everyone can have this experience:

If someone meditates for enough time, person becomes aware that he/she is not his/her body and can objectively be aware of impulses and information’s of his senses but knows it is not himself; it is not what he/she really is. Secondary one becomes aware of the flow of his emotions and able to understand mechanism of emotional reactions. Person can differentiate himself from his emotional being, control emotions without suppression into unconscious part of psyche.
Finally one becomes aware of his thoughts as a separate thing, becomes able to see their flow, as they emerge from nothing and then after few seconds disappear just to be replaced with other thoughts. Thoughts are like associative chains, coming after senses and emotions, as a reaction to them, or as a reaction to other thoughts, like chains. They can be restless and uncontrollable, or can bow to our will if we want to intellectualy comprehand something so we can solve a problem or get a better understanding of something.
So at that moment, if we are not thoughts, feelings, and our body, what we are? All is left is awareness itself, like some kind of presence without substance. This awareness has a lot of spiritual terms to be called, and can be experienced by everyone if they want to.
Unfortunately, on this level of scientific knowledge, we cannot explain it in a scientific way. There is no possibility to find a part of brain and show it, "yes, here it is, pure awareness, the very human soul", or do anything like that. We are unable to comprehend this in any way, because science is based on reason, on intellect, and experience can show that we are beyond intellect, and that human being just uses intellect as a tool, but is beyond that.

So even if we make machine that will have intellect, it will be just complicated program that will act like it is similar to human being, almost unable to distinguish from other human, but I doubt that it will have any kind of self consciousness, it will just pretend like having it, again based on programs given.