In a way, I do welcome our robot overlords. I suspect that robots – not having emotions and superstitions – have potential to be more moral than humans do. It is time for the human race to be self-modified or simply replaced by a better species.

Not that humans are bad… it’s just that we’re ancient technology. The world is being run on old Mac II computers. Sentient beings can do better.

The most important thing an AI computer could do would be to be able to break anything down to its most basic concepts, then reassemble them, with whatever differences it wanted. Moreover, it has to know the difference, which requires a sense of self, which is one thing we don’t know how to program into computers.

“Defining intelligence is nearly impossible. You said earlier that you don’t have a technical definition handy, but if you were to look for one you would find that a good one doesn’t really exist. Intelligence has a certain quality which is difficult to name, but which is recognizable when present.”

I find this statement curious indeed. Surely, we must be able to define it and measure it somehow. Isn’t that what our scientific tools are supposed to do for us?

“But if you believe that our brains are just complicated blobs of physical matter, with nothing supernatural (or quantum) happening inside of them, then we ourselves are just machines doing what we were programmed to do.”

Even if you believe that there are quantum processes going on, you can still make the same point. It’s only a matter of time before we design AI’s based on quantum computing.

“Finally, why should fooling a person into thinking it’s a human be the standard for AI? None of us would argue that a wolf (or almost any animal) lacks intelligence”

You’re right — I don’t think anyone would argue with this. We simply use human intelligence as a reference point, because that’s what we can relate to. Our efforts to create intelligence will (at least initially) likely be directed toward emulating human intelligence and measured against the same, because that is our most familiar measuring stick.

“Finally, my own opinion is that for intelligence to exist, it must evolve itself and it must have a “body” to live inside of. Whether the body exists only in a simulation or whether we make a robotic body that inhabits the physical world, having a sense of self (more or less by definition) requires some separateness from the environment”

That, my friend, is very astute. (OK, I’m not really your friend, but if McCain can say it all the time, can’t I? :-) I agree with you wholeheartedly.

“I guess you could argue that any program has the OS as its environment and its own memory space as its “body” but I think that would be stretching things considerably.”

Not really a stretch at all, IMO. Memory seems to be the key element in creating a sense of separateness. Just imagine for a minute what your experience would be like if you had none. I’d go a step further and say that ANY process, natural or manmade (of which a program is one example) involves an element of memory and intelligence.

Perhaps the reason we have a hard time creating an absolute definition of intelligence is because there really is none. If all processes do contain a measure of intelligence, then intelligence is not an threshold to be crossed, but a continuum that encompases everything we observe.

First, we already have “simple” AIs. Several of these have been mentioned already, and they do learn, solve problems on their own, etc. They may not conform to YOUR definition of intelligence but based on some of your other comments here I’m going to go out on a limb and say that you’re not an expert on defining intelligence or AI research ;)

Defining intelligence is nearly impossible. You said earlier that you don’t have a technical definition handy, but if you were to look for one you would find that a good one doesn’t really exist. Intelligence has a certain quality which is difficult to name, but which is recognizable when present. One of the biggest problems in attempting to define intelligence is that any time we name something that we think is intelligent, someone will present an example of a system that does exactly that but which doesn’t match our idea of what intelligence is despite meeting our requirements. I could give you examples of programs which are fully capable of learning and solving problems and you would still (probably) not call them intelligent. So we have kind of a moving goalpost situation, just like creationists do with evolution – every time evidence is presented they change the requirements.

You said that it should “not just [be] a glorified slide rule”, but that in itself is problematic because there are some very good arguments in favor of the human brain being nothing but a glorified slide rule itself. Human thought and consciousness can be modeled (albeit crudely) as a system that does nothing but shuffle symbols around based on simple rules. (See Godel, Escher, Bach by Hofstadter – a book that you really should read if you’re going to have discussions about AI.)

Same thing with “programmed responses”. A common argument against AI is that a computer can never actually be creative – it can only do what it was programmed to do. But if you believe that our brains are just complicated blobs of physical matter, with nothing supernatural (or quantum) happening inside of them, then we ourselves are just machines doing what we were programmed to do.

Finally, why should fooling a person into thinking it’s a human be the standard for AI? None of us would argue that a wolf (or almost any animal) lacks intelligence, but no-one would ever confuse a wolf for a human. Intelligence can apply to any problem, and in my opinion our efforts toward AI should not be to create a machine that can chat about the weather in a convincingly human manner. There are better (though perhaps less pleasing to the masses) things we could be doing with our time.

Finally, my own opinion is that for intelligence to exist, it must evolve itself and it must have a “body” to live inside of. Whether the body exists only in a simulation or whether we make a robotic body that inhabits the physical world, having a sense of self (more or less by definition) requires some separateness from the environment, and an ability to perceive the environment and act within it. I guess you could argue that any program has the OS as its environment and its own memory space as its “body” but I think that would be stretching things considerably.

Is it ever possible to prove that there is something we can’t and won’t ever explain through science? If so, what would be the criteria that frame that proof? If not, explain your reasoning.

What other forms of evidence do you propose that are rational?

When you build a house, you make it out of bricks or wood – real things. You build them on real foundations. You don’t build it in midair, because only other real things will support them. Is it then illogical to say that the only thing that can support real things is other real things?

We live in a world made of material things. Everything we can observe has an effect on other things. Moreover, if they don’t have an effect, they might as well not exist, because for all intents and purposes they don’t.

Moreover, what non-material things would you include in a rational system? Is there any logical reason for believing they exist at all?

Empirical evidence is the only rational criterion for understanding the world because… it is not rational to believe what we have no empirical evidence for.

Maybe I’m wrong, and that isn’t circular. I would appreciate being shown how, since I can’t seem to figure it out.

So if you think materialism is the only rational criterion for understanding the world, that’s cool by me. The only problem I have with it is that all too often people do not acknowledge the precarious logical leap that it takes.

“Johnny, I’d agree that the only meaning events have is what we give them… but I don’t see how it supports your point.”

Hmmm. OK. Might have been a stretch. I guess I was just saying that sometimes what seems like a constant is not really a constant — it only seems that way relative to our perspective, because of a certain meaning we or society assigns.

“Also, you seem to be reaching trying to disprove some of the constants brought up. Other universes might have different constants, but there’s no proof for other universes, so those physical constants are constant as far as we can see.”

Well, I had to try, didn’t I? :-) I’m not saying there’s proof here. But there is a lot of compelling evidence that it is possible — enough that we should keep looking to determine one way or another.

“The solipsistic argument is useless. Look at it this way. If there’s no evidence that the world is a simulation, it can either be real, or an incredibly good simulation. If it’s a simulation so good no test can tell the difference, ever, how can it possibly have an effect on our lives either way?”

I’m not sure solipsism is the right term here (the theory that I’m the only real consciousness, and the rest of you are just an illusion). I DO think that separateness is an illusion — but not in that way, and anyway, is going off topic… I do agree that the actual implementation of our universe may not make much difference to us — unless of course, we can learn to go beyond it. (Consider “The Matrix” as an example).

“You scorn materialism as “just another box”. But materialism isn’t a belief system, unlike what you seem to have adopted. It’s just a tool for testing the world. You put in observations, and what comes out is the closest you can get at that time to a view of the world around you.”

I didn’t mean to scorn materialism or anything else. I apologize if I came across that way. I just like trying to get people to think outside the box, and maybe get too pushy sometimes. I won’t deny having a system of beliefs — that’s a hard one to avoid — but I do try to remain aware of that fact and not let it prevent me from keeping an open mind. I’m perfectly happy redefining my “box” as it becomes appropriate to do so.

“If and when evidence comes up that the world is not real, that we are all just God’s Debris* or dreams in the mind of the creator, then materialism will lead those of us who use it to that conclusion and we will be at the forefront of figuring out the new limits on our existence. Because materialists don’t make something up and call it truth; they look at the world around them, see what it’s like, and call it truth for now.”

Fair enough. I certainly haven’t dismissed the idea that science/materialism can lead us to know everything there is to be known. I personally feel there are other tools at our disposal — but I understand that these tend to get very subjective.

“*Yes, I’ve read that book. Most of the science in it is bunk, but it’s fairly well written and a good story. I’ll definitely read the Religion Wars when I can get my hands on a copy.”

Cool — yeah, the science is mostly bunk — some of that is apparently intentional. I’m still not exactly sure what Adams’ point was in saying that there were ‘logical inconsistencies’ in the story — ‘can you find them?’ I’m guessing — and I could be wrong — that he wanted to say something about his own beliefs without being able to be pinned to them. Sort of a built-in escape clause.

And of course, as a philosophical theory, it’s a tool, just the same as Consequentialism, Feminism, Marxism, Postmodernism (the complicated, impossible-to-decipher and more or less useless thingamajig that your father picked up during his wandering years in the 60s of philosophical tools) and all the other isms. It just happens to be the ism that powers science, and therefore the most useful philosophical tool.

As for being rationalist without being materialist… how would you define rationality? To me, one of the most important parts of being rational is working from the evidence you have… which is basically materialism. If you’re just going to make things up and insist they have real substance (like God), rather than making things up as a way of inspecting the relationship between things (as we do in thought experiments), you’re not being rational anymore. At least, not by my definition.

I don’t know that you can call materialism nothing more than a tool.

dictionary.com gave me this:

1. preoccupation with or emphasis on material objects, comforts, and considerations, with a disinterest in or rejection of spiritual, intellectual, or cultural values.
2. the philosophical theory that regards matter and its motions as constituting the universe, and all phenomena, including those of mind, as due to material agencies.

Either directly opposed to certain ideologies (which puts it awfully close to an ideology itself) or a philosophical theory.

Do you think it is impossible to behave rationally without being a materialist?