50 replies to this topic

[arithma]

A quick and (I hope) specific question. In what ways you as a game programmer think it's possible that your created AI can realize your existence as its creator. Is it even possible for it to realize your existence. How does that apply to us and God?

[IADaveMark]

Dude... we can't even get it to realize 90% of its environment. You want us to get it to realize something that is NOT in its environment? What an odd question.

[ToohrVyk]

I have not yet heard of an artificial intelligence which can handle the concept of existence and creation in the same way humans do, so I would say that it's pretty simple to create an artificial intelligence which knows the existence of its creator (In Prolog: exists(creator).), but it's very difficult to give such concepts an observable meaning which we can deem correct enough.

As for God, not everyone is blessed with the awareness of His Noodliness, sadly.

[darookie]

According to Roger Penrose (see for example Shadows Of The Mind), consciousness is non-computable and cannot be modelled using computer programs. I tend to agree with this.

[Vorpy]

Of course a lot of scientists disagree with Penrose's conclusions. For example, the argument outlined in that wiki article is probably incorrect. A strong case can be made that there are limits to what the human mind can compute, and if that is the case then the entire godel argument collapses.

[apollodude217]

I think our creation of AI agents can be used as an illustration for our relationship to God in various ways. How? I'll leave that one up to you.

[cossie]

Quote:

How does that apply to us and God?

That's assuming there is a God ;-)

[DvDmanDT]

I'm still waiting for bluebrain.. :p Unless I missunderstood the entire project, it should be able to realize our existence, assuming the project succeeds.

[Palidine]

We can't even "prove" that a human who is not ourself is aware of their creator. How can I tell that they are not just an unfeeling robot programmed to respond "yes" to the question "are you aware of your creator"?

The closest I can get to proof is to say that I am aware of my creator, the other person is human so they probably think like me. Can you "prove" that? No. Can you "prove" it for a machine? No. Can you even "prove" that you, yourself, are aware of your creator? No.

Does science have a definition of consciousness? No. Therefore it's impossible to demonstrate that anything non-human possesses it. (non-human) because part of any potential definition of consciousness is "something humans have".

-me

[pinacolada]

while(True):
  q = raw_input("What is your question? ")
  if q == "Do you recognize me as your creator?":
    print "Yes"
  else:
    print "Huh?"

[Timkin]

I can only suggest that you all read the works of Gregory Chaitin. Start with his popular books and then get into his research literature. I had the pleasure and privelege of meeting and talking to him extensively when he visited my department a few years back; very exciting times! His research and ideas are certainly as important as Godel's and will go down in mathematical history as one of the greatest advances of modern mathematics.

For those not familiar with his work, he, along with Leibniz, is considered the founder of modern algorithmic information theory. He was the discoverer of Omega, being the halting probability of any program (a number that we can write down, but never compute!) When you get to the heart of his work, I believe that it leads to the conclusion that you certainly can reason with information using non-halting programs... but the result is that you will always have uncertainty...

Furthermore, one could take the position that internal to a non-halting program there is a usable information state that the program may itself use and manipulate to modify its behaviour, but it may never output that state (or else it could be considered to be a shorter, halting program). Thus, our brains+bodies may indeed be running a non-halting program, modifying their own computation, without having to produce an output (i.e., terminate and provide some 'result' of the algorithm+input).

[IADaveMark]

Of course, I'm thinking of Ender Wiggin's "Jane" a la Orson Scott Card.

[Kirl]

I do not have any authority on the subject of AI, other then that I am currently writting a cpu controlled snake wich still tends to crash into itself rather often (it, as well as myself still have a lot to learn). But I think it's an interesting subject.

I don't think AI programming can ever be compared to 'real' intelligence; our programming languages are all based on mathematics, and there is (currently anyway) no mathematical system that is capable of describing something as complex as intelligence. So I find it unlikely that we can ever create 'real' intelligence with it. I'm sure we can come a long way to faking it, but deep down they're all still based on hard (yes/no) rules, even if those rules incorporate some random factors.

I am however convinced that we will some day be capable of creating some form of intellingence, but I don't think with current computers and programming languages. Meaby when quantum computing becomes a reality...?

I am defenatly going to check out some of those books mentioned!



As for the original question, self-awareness is still the main problem. This question is not even in sight, interesting though.

Meaby interesting to read: http://www.pandasthumb.org/archives/2006/06/evolution_of_co_2.html

[PsyFly]

Here are some of my thoughts on it, on the philosophical side (since really this is a philosophical question)

Concerning God:
* human ideas of god, and to a great extent all spiritual and religious concepts, more or less stem from an interpretation of experiences or moments in which your consciousness is not functioning in a rational sense
* rather, most spiritual experiences involve a perspective focused on either a sense of awe and wonder, a loss of self, a feeling of union, etc.; mystic experience, peak experiences, and so on
* the concept of God is one way to interpret such an incomprehensible experience

So as far as our relationship to god, you could also ask what about maya, samsara, or simply consciousness and the universe; god is one idea resulting from an analysis of this intangible, selfless experience.

Concerning awareness of a creator:
* That would require the creature's cognition of time, self, and other oddities of OUR life. Appearing conscious and being conscious are also two very different things
* indeed, a theological argument could probably stand upon AI to attempt to rationalize God and creation (by comparing our minds to programs), but few theologians would willingly compare humans to machines, since that would imply a mechanistic, deterministic philosophy, something most religions argue against (instead arguing for "free will").

[MrRage]

Quote:

Original post by pinacolada while(True): q = raw_input("What is your question? ") if q == "Do you recognize me as your creator?": print "Yes" else: print "Huh?"

lol

But in all honesty you could probably make an AI that would simulate knowing its an AI, but it would just be a simulation. To make an AI with a personality and capable of independent thought, who knows.

[nashesolnce]

IF (We are the creators of the pc/internet-world){
IF(EveryCreator equal God) {
IF (pc/internet-world is more virtually, than you want){
We are the God of virtual pc/internet-world
}ELSE{
We are the God of real pc/internet-world
}
}ELSEIF( OnlyOneCreator equal God){
IF (CurrentLocalGod equal God)
We are tools in CurrentLocalGod's hands
}ELSE{
We are tools in unknown God's hands
}
}ELSE{
IF(rand(1,2)>1){
We all are Gods too...
}ELSE{
We all are tools...
}
}
}
)))))))

[Edited by - nashesolnce on April 28, 2007 10:35:26 AM]

[Hollower]

Umm, can I buy some pot from you? [wow]

[Timkin]

Quote:

Original post by Kirl I don't think AI programming can ever be compared to 'real' intelligence; our programming languages are all based on mathematics, and there is (currently anyway) no mathematical system that is capable of describing something as complex as intelligence.

Your assumption that 'intelligence' is complex is a common anthropomorphic assessment we humans tend to use to make ourselves feel superior to animals and insects and one that I don't think is justified. Much of what we do as humans (our behaviours) is biology and not 'intelligence'.

Quote:

So I find it unlikely that we can ever create 'real' intelligence with it.

Define "real intelligence" please.

;)

Cheers,

Timkin

[KreK]

Kirl, you do know that mathematics and physics are nothing more than languages used to describe the universe in which we live in? The basic premise of science is that everything that exists must follow natural laws and therefore can also be explained. Yes, even things such as love, hate and intelligence.

To claim that such things cannot be explained requires giving them some kind of supernatural and therefore unprovable, unscientific characteristics. Just because we don't yet have complete understanding of certain things doesn't mean we have to fall back to the same old and repeatedly disproven "God of the gaps" (better known as argument from ignorance).

I think the problem with "real intelligence" is going to be about its definition and this is something that people will probably never come to agree. What does "real" actually mean? Something that simply exists? Then we already have it in our pocket calculators. Perhaps you thought something that is human level? OK, then animals such as monkeys don't have a real but unreal intelligence.

From speculative fiction, does HAL9000 have real intelligence? Does commander Data have real intelligence or is he just a calculator? Do replicants from Bladerunner movie have real intelligence - they were after all just artificial humans and some would conclude that being artificial they cannot possibly have real intelligence. What's the difference between real intelligence and a perfect simulation of real intelligence?

No matter how you define "real" intelligence, you are going to crush some people world views and they will object to that definition.

[Edited by - KreK on April 30, 2007 6:45:52 AM]

[arithma]

In my mind while asking the question, I had already put it inside that humans are like machines: deterministic.
I was actually thinking about my own religious beliefs and comparing to the real-life things.
The issue I was pondering about is a model of communication between a creator and creation (with varying level of creating-ness, of God-Us larger than Us-AI).
I still think my question is a bit fuzzy, so I'll have to work on it more after I finish my finals, and maybe something cool will come out in the end (maybe I'll end an atheist, agnostic, or a highly religious dude-and with reasons to back it up :D).