Artificial Intelligence - Possible or Probable?

[jadler]

Hey everyone,

I am a CS major and have encountered various texts on the idea of "Turing" child machines and Artificial Intelligence. My question is - If someone were to state that they have created such a device, one capable of or with sufficient perceived capability of independent thought, would you believe that it could exist?

I ask because I am fascinated by the subject and am working towards a greater understanding of it's practical concepts. Within a few years of graduation I hope to join one of the projects in this field or start one of my own.

[Mr. Slant]

Your question is not amenable to "General Questions" as there's no factual answer.
You're asking multiple persons to express an opinion, which makes this an IMHO.

That being said, congrats on having discovered the wonders of AI.
Enjoy the field, and for Christ's sake please don't invent Skynet.

[Superfluous Parentheses]

If someone were to state that they have created such a device, one capable of or with sufficient perceived capability of independent thought, would you believe that it could exist?
No need to believe people at their word. We've got tests for this. I would be convinced if the machine repeatedly succeeds at a general turing test, but it won't be THAT hard to make a believable argument/test for intelligence even if it can't (for whatever reason).

[Q.E.D.]

Sure, it's absolutely possible. Our brains themselves are nothing but natural electrochemical Turing machines. That said, no, I wouldn't believe without strong evidence that someone today had created such a machine; we're still orders of magnitudes of complexity from even the stupidest vertebrate brain, let alone one equivalent to ours.

[Chronos]

Of course, you also have to set a standard for what counts as "artificial intelligence". We don't yet have a machine that can consistently pass a Turing test, but we do have machines, for instance, which can deliver you detailed information on a broad selection of topics, on demand. Is Google intelligent? In some ways, it's a heck of a lot smarter than you or I.

[Sage Rat]

Same as QED.

Personally, I suspect that creating a humanistic AI will come less from programming a particularly novel data structure or set of algorithms, but rather be nurtured slowly over a few decades from something simple but huge like neural nodes. So while it will be interesting to create programmatic toys for the proto-AI to use, essentially you're just brute forcing the solution through an evolution simulation--maybe even including rival AIs.

And again, while still cool, most "AI" research is going to be pretty irrelevant to this accomplishment. They're working on figuring out adaptive algorithms that don't take up all the space, but most likely have a limit to how far they can get regardless of the amount of memory they're given. They'll only ever be able to adapt to a single task.

[levdrakon]

I wouldn't believe someone if they said they made a Turing-test-passing AI today.

First, I'd need to see the AI that can pass the insect test, then the cat test, then the dog test, then the chimpanzee test, then the Turing test. I don't think a human-like AI is going to take us by surprise; we'll all get to watch it develop.

[Exapno Mapcase]

I ask because I am fascinated by the subject and am working towards a greater understanding of it's practical concepts. Within a few years of graduation I hope to join one of the projects in this field or start one of my own.
Right now, the truly difficult questions in the field are in the province of biology rather than computer science. We know a great deal on the silicon side, and merely infinitesimal amounts on what we need to know on the carbon side.

We don't even have beginning working definitions of intelligence, let alone of how the brain functions.

If you want to get into the field, I'd suggest expanding your scope of study in graduate school to begin examining the biological side of the problem. That will give you a big leg up in the future.

[ultrafilter]

Sure, it's absolutely possible. Our brains themselves are nothing but natural electrochemical Turing machines.

This is probably true, but given how poorly we understand the functioning of the human brain, it's not reasonable to present it as accepted truth.

[Q.E.D.]

This is probably true, but given how poorly we understand the functioning of the human brain, it's not reasonable to present it as accepted truth.

Of course it is. What realistic alternative could you possibly conceive?

[ultrafilter]

Of course it is. What realistic alternative could you possibly conceive?

I can't think of anything else, but the fact that I can't imagine something doesn't mean it's possible.

[Q.E.D.]

I can't think of anything else...

Neither has anyone else which is why it is, in fact, reasonable for me to post what I did as accepted truth. Since it is.

[Absolute]

Of course it is. What realistic alternative could you possibly conceive?
I don't you're using the term "Turing Machine" correctly.

A Turing Machine is an abstract, theoretical computer design that can be shown to be equivalent to any other computer design (in terms of what calculations it can perform). This does not mean that all computers are Turing machines, and certainly not that the human brain is a Turing machine. For one thing, the hypothetical Turing machine has unlimited memory.

Specifically, a Turing machine has a linear memory consisting of symbols, a state register, a table of state transitions, etc. Unless you're arguing that there is a biochemical equivalent of these mechanisms in the human brain?

I think you are trying to say that the brain is merely a biochemical computer, and that it does not perform any operations beyond those that an (electro)mechanical computer can perform. While it is possible that the brain makes use of quantum effects, I would be inclined to agree with this opinion.

----

On the topic of Turing, the Turing test is actually a pretty poor way to test for actual intelligence. One can imagine a computer that is programmed to B.S. it's way out of any questions regarding learning, abstraction, and that sort of thing.

[Sage Rat]

On the topic of Turing, the Turing test is actually a pretty poor way to test for actual intelligence. One can imagine a computer that is programmed to B.S. it's way out of any questions regarding learning, abstraction, and that sort of thing.
I don't think so. The turing test assumes that you're specifically trying to prove that it's unable to perform human tasks, not just having a casual conversation. A BS machine wouldn't be able to learn a programming language and to code something up according to specification, for instance.

[Omphaloskeptic]

Sure, it's absolutely possible. Our brains themselves are nothing but natural electrochemical Turing machines.This is probably true, but given how poorly we understand the functioning of the human brain, it's not reasonable to present it as accepted truth.Of course it is. What realistic alternative could you possibly conceive?I can't think of anything else, but the fact that I can't imagine something doesn't mean it's possible.Neither has anyone else which is why it is, in fact, reasonable for me to post what I did as accepted truth. Since it is.That's a little too strong. Penrose, for example, has certainly claimed (The Emperor's New Mind, etc.) that the brain is stronger than a Turing machine, and even has a suggestion about how that might be possible. I don't agree with him (and his quantum-gravity idea is :dubious:), but there are some pretty well-known people who don't accept the strong-AI or weak-AI positions.

[DanBlather]

IOn the topic of Turing, the Turing test is actually a pretty poor way to test for actual intelligence. One can imagine a computer that is programmed to B.S. it's way out of any questions regarding learning, abstraction, and that sort of thing.Are you saying politicians are not intelligent?

[runcible spoon]

Politician Intelligence - Possible or Probable?

[Washoe]

That said, no, I wouldn't believe without strong evidence that someone today had created such a machine; we're still orders of magnitudes of complexity from even the stupidest vertebrate brain, let alone one equivalent to ours.

Really? If we analogized the most sophisticated computer currently in existence to an animal's brain, how would it stack up? Like an insect's? A mouse's? A dog's? Less than an insect's?

[Cerowyn]

The problem with trying to draw analogies between organic brains and AI systems is that they are only superficially similar: they take input, process it, and provide appropriate responses. Organic brains are massively parallel, performing many disparate functions simultaneously. A human does not consider walking to be an intensive operation, and yet it took a long time for processing and feedback systems to be able to replicate that process. But no one would ever suggest that they could perform mathematical operations at even the tiniest fraction of the speed that a home computer is capable of.

Saying that a modern AI system does not compare to the stupidest vertebrate brain is certainly true. From a certain perspective. Nearly ten thousand people a day in Japan talk over a phone to a system that my company wrote to ask questions about their bank. It understands a broad range of speech, and responds appropriately even when it doesn't have the answer. We've read bloggers theorizing that the system cheats by having human operators provide the responses interactively (which would sort of undermine the whole reason for having the software-based system!). Does that pass the Turing test*? No, not even remotely. It's not too difficult to trip the system up if you're trying, and you notice very quickly if you stray too far outside of the knowledge domain that has been created for the agent. Even our weak AI (http://en.wikipedia.org/wiki/Weak_AI) system is a long way from the standards of being true AI by any definition that most people would use.

* Not to be confused with a Turing machine, which is described by Absolute above.

[Half Man Half Wit]

I wouldn't believe someone if they said they made a Turing-test-passing AI today.
Well, they did get pretty close this year: the requirement is that 30% of investigators interrogating the machine have to be fooled into thinking it was in fact a human being, and 25% were (http://www.telegraph.co.uk/earth/main.jhtml?xml=/earth/2008/10/12/eacomputer112.xml). That's not insignificant.

Regarding the OP, I essentially share QED's views on the brain, and so I don't see any insurmountable obstacles on the path to true AI.

[Q.E.D.]

Really? If we analogized the most sophisticated computer currently in existence to an animal's brain, how would it stack up? Like an insect's? A mouse's? A dog's? Less than an insect's?

Yes, really. The best computers we have today can't compete with the intelligence of even a mouse. We're close to insect level. But, insects aren't vertebrates.

[CookingWithGas]

I graduated with a CS degree in 1979 and took a course in AI. Although I have not seriously studied the AI field since then, my casual observations lead me to think that there has been little progress since then, and certainly no breakthroughs, towards a man-made non-biological device with "perceived capability of independent thought."

[ultrafilter]

I graduated with a CS degree in 1979 and took a course in AI. Although I have not seriously studied the AI field since then, my casual observations lead me to think that there has been little progress since then, and certainly no breakthroughs, towards a man-made non-biological device with "perceived capability of independent thought."

There hasn't been much progress towards strong AI, but there's been a lot of work on intelligent systems and machine learning. Nowadays, intelligent systems are literally ubiquitous, and it's difficult to describe the impact of machine learning research without sounding like I'm exaggerating. Tom Mitchell has a nice whitepaper (http://www-2.cs.cmu.edu/~tom/pubs/MachineLearning.pdf) (pdf) that describes the field and some of its impact.

[levdrakon]

Well, they did get pretty close this year: the requirement is that 30% of investigators interrogating the machine have to be fooled into thinking it was in fact a human being, and 25% were (http://www.telegraph.co.uk/earth/main.jhtml?xml=/earth/2008/10/12/eacomputer112.xml). That's not insignificant.

Regarding the OP, I essentially share QED's views on the brain, and so I don't see any insurmountable obstacles on the path to true AI.True, I do believe a computer is going to pass the Turing test, but if it happened today I wouldn't believe it was doing anything more than brute-force guessing its way through billions of possibly correct answers, but not displaying true intelligence or understanding.

Like QED says, we can barely understand and duplicate an insect brain. Which is cool, but insect brains are on a different branch of the evolutionary tree. When we understand and can duplicate how the simplest mammalian brain works, we'll be on to something. I'd expect that would be in the news, and the sudden appearance of a truly intelligent, learning and understanding AI wouldn't just appear tomorrow.

Once we truly understand how a shrew or mouse brain works and can duplicate it, then I do think we can let the incredible speed and processing power of modern computers do in a short period of time, what evolution took 125 millions years to do, which is basically just trying out random mutations and keeping what works, towards the path to intelligence. Once we figure out that first mammalian brain, computers should be able to try out 125 million years of random mutations pretty quickly.

[ralph124c]

Why attempt to copy the human brain? Our airplanes do not fly like birds do. It would be much more productive to look at ways to complement human reasoning (expert systems). Technically, much easier, and potentially just as valuable.

[levdrakon]

Why attempt to copy the human brain? Our airplanes do not fly like birds do. It would be much more productive to look at ways to complement human reasoning (expert systems). Technically, much easier, and potentially just as valuable.More ethical too. I'd feel better about a super chess computer that could figure out novel solutions to our problems, without worrying if it were conscious.

[ultrafilter]

True, I do believe a computer is going to pass the Turing test, but if it happened today I wouldn't believe it was doing anything more than brute-force guessing its way through billions of possibly correct answers, but not displaying true intelligence or understanding.

Can you prove that a randomly chosen person displays true intelligence or understanding? It's remarkably more difficult than you might think.

[Half Man Half Wit]

True, I do believe a computer is going to pass the Turing test, but if it happened today I wouldn't believe it was doing anything more than brute-force guessing its way through billions of possibly correct answers, but not displaying true intelligence or understanding.
Well, then it's debatable whether there is something like 'true' intelligence or understanding, at least as determinable from the outside -- if nobody could say with better than chance success whether they are conversing with a human or a computer, then the two have to be equated in intelligence and understanding, at least when viewed from the outside. It might well be that the computer able to do so still isn't in any way conscious, but then, strictly speaking, the same qualification applies to your fellow humans, as well.

Once we truly understand how a shrew or mouse brain works and can duplicate it, then I do think we can let the incredible speed and processing power of modern computers do in a short period of time, what evolution took 125 millions years to do, which is basically just trying out random mutations and keeping what works, towards the path to intelligence. Once we figure out that first mammalian brain, computers should be able to try out 125 million years of random mutations pretty quickly.
Oh, I'm not sure we should view the mammal brain as all that special -- while it's demonstrably one structure that leads to the development of intelligence, that doesn't mean it's the only one. Indeed, if somebody came up with a giant spreadsheet that correlates every possible input with a reasonable output, in the 'outside view' definition, that construct would be intelligent, as well.

But this is all probably veering too far into GD territory.

[Chronos]

Yes, really. The best computers we have today can't compete with the intelligence of even a mouse. We're close to insect level. But, insects aren't vertebrates.Then again, a mouse can't compete with the intelligence of even a rudimentary computer, either. While I don't doubt that we'll eventually have a computer that can do everything a human can, there's very little incentive to push in that direction. If I want something that can carry on a conversation with a human, I can hire a human. But if I want something that can perform multi-dimensional nonanalytic integrals quickly and reliably, for instance, hiring a human to do it isn't really an option, so there is an incentive to get a computer capable of doing that.

[Stranger On A Train]

I don't you're using the term "Turing Machine" correctly.

A Turing Machine is an abstract, theoretical computer design that can be shown to be equivalent to any other computer design (in terms of what calculations it can perform). This does not mean that all computers are Turing machines, and certainly not that the human brain is a Turing machine. For one thing, the hypothetical Turing machine has unlimited memory.In fact, the human brain (or those of any other animal, for that matter) probably cannot even be considered as Turing complete except in a very restricted sense. This is not to say that the brain is not an incredible biochemical data processing and interpreting system, but it is fundamentally unlike the semiconductor microprocessor and support system sitting on your desk in many salient ways. Cerowyn details this extensively so I don't see any reason to reiterate, but it should suffice to say that we have only the most tenuous grasp on what is necessary and sufficient for human cognition, much less build a computer, operating system, and unities that could be indistinguishable from 'natural' human intelligence. The Turing test is really a very crude, somewhat subjective, and high level assessment of intelligence. Even if you could build a machine that can pass a Turing test based upon a statistical blind evaluation across a representative human population, it wouldn't prove that the machine is intelligence, only that it meets the criteria for Turing's evaluation of communication capability. On the other hand, a machine could have legitimate synthetic intelligence and be completely incapable of interacting on a natural language level.

When we do achieve genuine synthetic or artificial intelligence, I predict that both the hardware and software will be indistinguishable from the nervous system of a living organism in structure and operation, and it will look nothing like your Xbox.

Stranger

[RaftPeople]

Then again, a mouse can't compete with the intelligence of even a rudimentary computer, either.

In what ways do you think the intelligence of a mouse can't compete with a rudimentary computer? Are you thinking number crunching?

[Bad Astronaut]

That's a little too strong. Penrose, for example, has certainly claimed (The Emperor's New Mind, etc.) that the brain is stronger than a Turing machine, and even has a suggestion about how that might be possible. I don't agree with him (and his quantum-gravity idea is :dubious:), but there are some pretty well-known people who don't accept the strong-AI or weak-AI positions.
I think Turing machines were only mentioned with reference to passing a test for intelligence. The question of interest seems to be "can a machine be built that demonstrates human-like intelligence," and based on the existence of human brains I can't see how it would be reasonable to answer in the negative.

[Chronos]

In what ways do you think the intelligence of a mouse can't compete with a rudimentary computer? Are you thinking number crunching?Sure, why not?

[Stealth Potato]

In fact, the human brain (or those of any other animal, for that matter) probably cannot even be considered as Turing complete except in a very restricted sense. This is not to say that the brain is not an incredible biochemical data processing and interpreting system, but it is fundamentally unlike the semiconductor microprocessor and support system sitting on your desk in many salient ways.

Really? What restricted sense do you mean? I mean, modulo the same lack of infinite memory that we forgive in many other things we call Turing complete, it certainly seems to me that a human brain is, with time and effort, capable of performing a simulation of many Turing-equivalent systems. Conway's Game of Life springs to mind, for example.

I definitely get your point about the brain being fundamentally unlike electronic computing systems, but from a strictly theoretical standpoint, it seems to me fair to say that the brain is Turing complete. However, if you know of any good arguments on the matter I'd be very interested in them. I don't know of any academic investigations into the matter, and I didn't really look into the Turing completeness of the brain while I was in school, so I'm just going by my intuitive impression here.

[Absolute]

Really? What restricted sense do you mean? I mean, modulo the same lack of infinite memory that we forgive in many other things we call Turing complete, it certainly seems to me that a human brain is, with time and effort, capable of performing a simulation of many Turing-equivalent systems. Conway's Game of Life springs to mind, for example.

I definitely get your point about the brain being fundamentally unlike electronic computing systems, but from a strictly theoretical standpoint, it seems to me fair to say that the brain is Turing complete. However, if you know of any good arguments on the matter I'd be very interested in them. I don't know of any academic investigations into the matter, and I didn't really look into the Turing completeness of the brain while I was in school, so I'm just going by my intuitive impression here.
The brain is definitely Turing-complete. Given an infinite surface on which to scribble stuff, we can follow the rules that define a Turing machine.

So, we know the brain can do anything a Turing machine can do, albeit slowly and rather unreliably. The question is whether a Turing machine can do anything the brain can do.

[levdrakon]

Can you prove that a randomly chosen person displays true intelligence or understanding? It's remarkably more difficult than you might think.I'm not into this "outside view" thing. If it walks like a duck, and quacks like a duck, that's great in deductive reasoning, but not so great when I can create waddling, quacking robots.

[Half Man Half Wit]

I'm not into this "outside view" thing. If it walks like a duck, and quacks like a duck, that's great in deductive reasoning, but not so great when I can create waddling, quacking robots.
Well, it'd be easy enough to cut open one of your robots to see that its insides aren't that ducklike at all, but how do you tell a machine -- or human, for that matter -- that seems intelligent (through producing apparently intelligent responses to outside stimuli) from one that actually is intelligent?

[levdrakon]

Well, it'd be easy enough to cut open one of your robots to see that its insides aren't that ducklike at all, but how do you tell a machine -- or human, for that matter -- that seems intelligent (through producing apparently intelligent responses to outside stimuli) from one that actually is intelligent?Well, couldn't we also look at this computer's programming, and see how it's going about its answers? What would be interesting is if we looked at its programming and discovered it had "evolved" its programming in ways we no longer understand.

[Chronos]

Well, couldn't we also look at this computer's programming, and see how it's going about its answers?Sure, we can see what the computer's programming looks like. But that's not enough. If the question is "does the computer work the same way as a human?", then we also need to see the human's source code to be able to compare them.

[YamatoTwinkie]

Sure, why not?

But number crunching isn't intelligence. I can multiply two big numbers on a calculator quicker than I can in my head (and much quicker than a mouse could), but the calculator itself isn't any more intelligent than a stack of paper multiplication tables is. The calculator doesn't grasp the concept of numbers, or know what multiplication actually means.

A mouse, on the other hand, can grasp concepts (albeit basic), and can truely recognize patterns. It's vastly more intelligent than any supercomputer ever created.