Is consciousness necessary for intelligence?

[Guv]

Most people have an understanding of what intelligence is, or at least agree on what intelligent behavior is. While everybody might not agree 100% on what constitutes intelligent behavior, I assume that we do not need a definition for intelligence because most views of it overlap to a large degree. Humans have it, rocks do not, & there is a lot of controversy over where to draw the line between all entities or objects which have intelligence and all that do not. If you need to discuss or define what intelligence is, please start another Thread.

Consciousness is a bit trickier. It is often described or defined as awareness of self. I like to think that my life experiences are like a movie being watched or a book being read, with my consciousness being the viewer or the reader.

When I ski, play tennis, play chess, pursue a female, write a program, solve a mathematical problem, et cetera, there are instinctive activities and intelligent activities involved in accomplishing what I am doing. There is also an entity (my consciousness) which has thoughts like: I am enjoying what I am doing, or I wish I were doing something else, or I am doing this well/poorly. I hope we do not need more discussion of what consciousness is.

When Deep Blue was playing chess versus Gary Kasporov, there was no part of the system that was viewing the movie or reading the book about playing Kasporov.

Most people would consider chess playing as intelligent behavior. Yet Deep Blue shows that it can be done without consciousness.

[SteveCRM]

I think for intelligence such as Big Blue's, consciousness is not needed. Big Blues intelligence was just find the most logical and the move that got the most points. There was no creativity or planning. It was run on facts. A conscious intelligence in my view is one that has creativity. You mentioned saying "I am enjoying what I am doing, or I wish I were doing something else, or I am doing this well/poorly." this tells you if another approach to a certain problem is needed. A being with intelligence and consciousness is able to think through certain problems, and think ahead, while an intelligence without consciousness would rely only on facts it sees or is given.

Any by the way, when did deep blue play this game? (im 15, not sure how long ago that was) And by your words it sounded like he won.

[Guv]

Deep Blue defeated Gary Kasporov about 4-6 years ago. I think it was in 1997, but am not sure.

They played a series of games, as is the case when two humans play for the world title. I do not remember how many games they played nor the final score. Kasporov won at least one game, but there was not doubt that the computer won easily.

Kasporov said he was surprised at how sophisticated a player the computer seemed to be. He had expected play that would be somehow more mechanical or unimaginative. He expected to be able to find some weakness peculiar to a computer which he could exploit. Earlier chess playing programs had strange flaws which could be detected and exploited.

It is interesting that the programmers, individually or as a group, could not come close to beating Kasporov.

It is also interesting that Deep Blue is not considered an example of AI by anyone who understands the basic nature of the program.

It is basically a number cruncher. Board evaluation functions and Von Neuman Minimax strategy are the basis for most of the program's activities. These are augmented by a data base of book openings and various programs used to handle special situations.

[simonm]

Guv

OK, you are adamant that you don't want to discuss the definition of intelligence and seem reluctant to discuss the definition of consiousness and I understand why you don't want to get dragged down these side tracks. I will therefore leave the debate about "What is intelligence" for elsewhere but I must clarify your definition of 'Consiousness'.

Consciousness is a bit trickier. It is often described or defined as awareness of self.

Consiousness is Awareness. Self-consiousness is Self-awareness.

So is the point of this thread to assertain whether self-consiousness is a pre-requisite for intelligence?

I would define consiousness as the sensation of experience.

Kasporov said he was surprised at how sophisticated a player the computer seemed to be. He had expected play that would be somehow more mechanical or unimaginative. He expected to be able to find some weakness peculiar to a computer which he could exploit. Earlier chess playing programs had strange flaws which could be detected and exploited.

I heard that later, Kasporov was angry because he claimed that the computer had tricked him by luring him into a trap. The point is that the programmers had laid the trap and not the computer. As you stated (elsewhere), the computer is not intelligent attall. It just crunches numbers and follows a program.

I would say that self-consiousness is required for intelligence beyond a certain degree. Take a fox, for example. A fox can be very clever in the way it hunts and stalks it's prey and exhibits, most would agree, a fair degree of intelligence.
But, I think there is only so far you can go without self-awareness, without a comprehension of the relationship between one's self and the universe around it.

In order to hypothesise about one's self in imaginary situations it requires an awareness of the self.

I have a theory (that I developed based on reading books by Edward De Bono) that there are basically two things in the brain competing for your attention: Your senses and your internal brain patterns. Animal attentions are almost completely directed by their senses (and hence they have better sences than we do) and Human brains are more evenly balanced between the two. The greater the shift from the senses to the internal brain patterns results in less rigid brain patterns, greater detachment from reality and a greater capacity for intelligence.

[HarryW]

First I'd like to say that if we are going to discuss consciousness then we should do it in another thread so as not to fill this one with stuff. Secondly I'm going to ignore my own advice and respond to Simon's post.

I disagree with your definitions of consciousness. Awareness is not enough for consciousness, awareness just implies a gathering of knowledge about the environment. Perception of the environment is a prerequisite for intelligence (with no input there can be no cognition) so by this definition all intelligent entities are conscious. Consciousness is generally thought to be much more complex than that, and although I'm not entirely comfortable with defining it simply as 'self-awareness', I am much happier with that than simply 'awareness'.

[Geoff Gunson]

I have always though that a self aware computer would be impossible, however my old man said something quite profound the other day (he's not computer literate) (i can't spell). He said that it is almost impossible to prove an impossiblity and that AI might happen by mistake. For Instance the bods designing some new organic computer not aiming for AI accidently build it.

I like that thoery better than actually trying to create AI because it could form like life itself, just kinda happning.

[HarryW]

It did take millions of years for it to just happen the first time though, so I don't think we should rely on that method if we want to develop some form of conscious AI

[simonm]

I disagree with your definitions of consciousness. Awareness is not enough for consciousness, awareness just implies a gathering of knowledge about the environment.

OK then, let's define awareness (as far as I know, no one's ruled out discussing that yet).

A computer gathers information through input devices such as a keyboard and a mouse. If your definition of awareness includes a computer by means of it's keyboard and mouse then I would say that your definition of awareness is different than mine.

i.e. A computer is not aware, even though it gathers information through various input devices.

I did go on to my specifically define consiousness/awareness as "the sensation of experience". This, I believe, more acurately describes my definition of consiousness.

Consciousness is generally thought to be much more complex than that, and although I'm not entirely comfortable with defining it simply as 'self-awareness', I am much happier with that than simply 'awareness'.

I think self-consiousness is distinct from consiousness. With the above definition, they would be the same? Or would Self-consiousness = Self-self-awareness?

It's all very well saying the definition of 'Consiousness' should be debated in another thread but the distinction between consiousness and self-consiousness is too great (in my mind) to be able to continue debating this thread without resolution (one way or the other).

[HarryW]

Your average bog standard computer that Joe Bloggs at home runs Word on (and was probably convinced he needed a P3 1GHz for the privelege, but that's another argument ) has only a very limited range of input options, usually keyboard and mouse, maybe a microphone or a webcam too, but we aren't talking about basic computers here, we're talking about AI. Current AI is capable of analysing its environment in great detail, including mapping out the area it resides in within visual range, analysing and to some extent understanding speech, even breaking down and recognising smells. That is not a trivial amount of perception, and to me that is an awareness of its environment, although it doesn't mean that these AIs are conscious.

[simonm]

Phew! Finnaly able to post again after not being ablt to get on the site since yesterday.

Your average bog standard computer that Joe Bloggs at home runs Word on (and was probably convinced he needed a P3 1GHz for the privelege, but that's another argument ) has only a very limited range of input options, usually keyboard and mouse, maybe a microphone or a webcam too, but we aren't talking about basic computers here, we're talking about AI.

Well, I was using the analogy of a standard computer to demonstrate that the mere gathering of knowlege about your environment is not enough to make you aware (no matter how sophisticated your information gathering devices are).

Current AI is capable of analysing its environment in great detail, including mapping out the area it resides in within visual range, analysing and to some extent understanding speech, even breaking down and recognising smells. That is not a trivial amount of perception, and to me that is an awareness of its environment, although it doesn't mean that these AIs are conscious.

OK, if I could detach my eyes and plug them into my PC through an interface device so that the computer could record information comming through the eye, would my PC now be considered 'aware'? I think not.

Until a computer can obtain the "sensation of experience" then it is not really aware.

[HarryW]

I think you're missing the point of what I'm saying, and I also think we have different views of what it is to be aware. It's not the sophistication of the camera, or the sensitivity of the microphone, or anything else that I'm trying to emphasise, but the cognition of the information made available through them. You might be able to 'plug in your eyeballs' but that wouldn't make a great deal of difference, because it's not just about gathering data, it's about processing that data, assimilating it into information and, from that, asserting knowledge of the environment.

The AI may not be aware that it's doing that, but then it's not self-aware is it?

[simonm]

OK, I agree, we probably do have different definitions of "aware". But you were the one that said:

...awareness just implies a gathering of knowledge about the environment.

For me, the above definition is inadequate to encapsulate what I think of as awareness. For me, something is not aware unless it has the sensation of experience. As far as I know, there has not been a computer developed that has that yet because scientists still cannot agree on how that arises in the brain (any brain let alone humans).

The next step from awareness to self-awareness is merely an extension of the above definition. To be self aware is to make the intuitive leap to grasp the concept of self. It is an intuitive leap because the concept of self is an infintely recursive thought. This comprehension of oneself gives us the ability to project ourselves into imaginary situations in order to postulate an outcome.

The advantages of self-awareness to the development of intelligence are plainly self evident.

[HarryW]

'Sensation of experience' is another pharase like 'framework of paradigms', it's kinda fuzzy and not very expressive.

People have been moving the goalposts of what would constitute 'true' AI for years now. It's been said in the past that we would have achieved 'true' AI when a computer could play chess and beat a grandmaster, and when a computer could understand a conversation and answer questions on it later, and other supposed landmarks. These have been achieved only to have us decide that they were missing that quintessential something that we couldn't quite put our finger on yet, but was definitely not there, so we make up new targets that we'll ignore when they get achieved.

I doubt you will ever be able to convince yourself that an AI has 'sensation of experience' in the same way you do because artificial intelligence seems so far removed from our intelligence. I'm not suggesting that AI has come that far yet obviously.

I seem to remember there was one computer scientist, active in the field of AI research, who was convinced that his laptop was conscious. Perhaps he was right.

[simonm]

'Sensation of experience' is another pharase like 'framework of paradigms', it's kinda fuzzy and not very expressive.

Well I appreciate the fact that they do not convey as much meaning as they should although they are as close as I can get (which ain't that close) to expressing what I'm trying to say.

As far as the goal posts having moved over the years with relation to the development of AI, then you're probably right but I think this has mainly been brought about through our improved understanding of human intelligence and brain functioning. The more we learn about the brain, the more we realise how far short of actual intelligence computers are.

But, you do have a point. Will it ever really be possible to determine whether something is actual aware/consious or not? Can you prove you're consious? How do I know you aren't just processing symbols mechanically? Do you really understand the meaning of the symbols? We can prove to ourselves we are consious only because we are self-consous. A computer, therefore, would not be able to prove to itself that it is consious until it is self-consious (and then not to anybody else).

If someone could come up with a test that determined whether or not something was consious or not then that would be helpful. I think we already have a test for the ability to be self-consious; From Kurt Godel I believe.

[HarryW]

Well your 'self-conscious' is my 'conscious' so I can't really apply the same logic that you are using.

The other points are taken, although I think intelligence isn't too hard to achieve. Perhaps it's not as far developed as human intelligence (well there's no 'perhaps' about it) but it is basically intelligence nonetheless.

[simonm]

Well your 'self-conscious' is my 'conscious' so I can't really apply the same logic that you are using.

OK, if you mean 'Consious' when I mean 'Self-Consious' then what do you mean when you say 'Self-Consious'.

The other points are taken, although I think intelligence isn't too hard to achieve. Perhaps it's not as far developed as human intelligence (well there's no 'perhaps' about it) but it is basically intelligence nonetheless.

Have you read Richard Penrose's "Shadows of the Mind" and "Emporer's new mind"? He makes a very strong case for the incomputability of human intelligence.

[HarryW]

Funnily enough... no I haven't, I've never heard of them Care to give a (very) brief overview?

'Self-conscious', to me, isn't really on the same level in terms of importance as 'conscious'. It means something more like being particularly attentive to your current state and behaviour, ie more so than usual. It's more like a colloquial term.

[simonm]

'Self-conscious', to me, isn't really on the same level in terms of importance as 'conscious'. It means something more like being particularly attentive to your current state and behaviour, ie more so than usual. It's more like a colloquial term.

Yes, in every day language, it does mean to be overtly consoius of one's own appearance.

But, technically speaking, to be consious is to be aware. You can be consious of someone's presence, of the light being on, of the fact that it is raining etc. The point being that the "self" is jost one thing to be consious of.

Funnily enough... no I haven't, I've never heard of them Care to give a (very) brief overview?

OK, brief it is...

Basically, his main argument against the computability of human intelligence rests on his interpretation of Godel's incompleteness theorems. He looks at many other aspects of the human mind including "free-will" which he uses quantum mechanics to propose a non-deterministic process by which thoughts can come about.

Perhaps the future lies with Quantum computers. Maybe there is only so far we can go with conventional computation systems. Perhaps, it is not necessary to go "all the way" and we can get to a sufficient level of "intelligence" for our needs.

[HarryW]

If you believe in quantum randomness, as far as I know a pretty big part of quantum mechanics, then nothing is deterministic, not just thought processes. The subject of free will to me is dead and buried. People place far too much importance on it just because they're uncomfortable with the idea they might not have it.

So, the question is: if human intelligence is incomputable, how does the human brain work its magic?

[simonm]

The subject of free will to me is dead and buried.

Perhaps as far as you're concerned but it's hardly been resolved satisfactorally one way or the other yet. Still that's another side road I'm not going to go down now.

So, the question is: if human intelligence is incomputable, how does the human brain work its magic?

Well, first I would like to say that it is not necessarilly my opinion, I was just expressing the counter arguments.

I have read a scientific paper on "consiousness" (sorry, I can't remember by whom as it was quite a while ago now) and the author made a strong case for showing that consiousness itself could not possibly arise from a mechanistic process. i.e. It is fundamentally non-mechanistic in nature.

So what does that leave us with then? Something not very satisfactory (to me anyway). Something mystical like a soul, spirit or higher plane of existance.

I should say that he didn't attempt to prove any particular "mystical" means, rather he concentrated on demonstrating how consiousness could never result from physical processes in the brain.

I personnally don't really subscribe to this "mystical" point of view but until we have come up with a satisfactory test to determine whether something is consious or not then we are going to be stuck (because we will never know for sure whether we have achieved it or not).

[HarryW]

I didn't say I thought I knew whether free will existed or not, I just said the subject was dead and buried to me. As in pointless. It's irrelevant. Doesn't matter either way.

The fact is human-level intelligence is possible, that is evident from the fact that humans have it. Whatever it arises from is either a product of the laws of physics, in which case it is repeatable (although possibly repeating it is not within the capabilities of the human race) or a product of some divine/magical intervention, in which case there will be a lot of rather surprised people, and a lot of rather smug people.

Anyway my bet is, I have to say, on the former.

[simonm]

Here is a website that explores mankinds search for an understanding and a definition of consiousness:

http://dubinserver.colorado.edu/prj/kmu2/2.html

It goes on to examine the debate of Physicalism vs. Dualism.

I haven't read it all yet but it is quite interesting.

[Kzin]

Originally posted by HarryW
I didn't say I thought I knew whether free will existed or not, I just said the subject was dead and buried to me. As in pointless. It's irrelevant. Doesn't matter either way.

The fact is human-level intelligence is possible, that is evident from the fact that humans have it. Whatever it arises from is either a product of the laws of physics, in which case it is repeatable (although possibly repeating it is not within the capabilities of the human race) or a product of some divine/magical intervention, in which case there will be a lot of rather surprised people, and a lot of rather smug people.

Anyway my bet is, I have to say, on the former.

Harry I like your line of reasoning in this thread - but the problem discussed here basically pivots on each person's interpretation of Searle's Chinese Box (or Room) Paradox.

In summary - if something behaves as though it's conscious then is it? If it is not then conscieness is essentially tied to some arbitary definition - e.g. it's a eukaryote, a mammal, a human, my race, my nation, just me or some other arbitrary chauvinistic claim.

I'd very much like to hear Guv's view on Searle too!

[Guv]

I believe that, in principle, intelligence can exist without consciousness. In practice, I am not so sure it ever will.

Some intelligent people have come up with arguments claiming that only a biochemical entity can have consciousness. Equally bright people argue that an electronic device could have consciousness. I side with the latter.

A few posts seem to imply that AI is far more advanced than I think it is.

I am glad the thread has not digressed into discussion about the definition of intelligence.

It may seem that the AI goal posts are constantly moved back, but I do not think that this is a fair assessment of what is happening. When some very bright people claimed that a computer beating the best human chess player would be an example of AI, they did not think it could be done by number crunching and the Minimax algorithm. They thought that it would require something like the human brain to do it.

BTW: Maybe if we knew how humans played chess, perhaps we would say something like: “That doesn’t seem to require intelligence. It is just clever pattern recognition and the use of memory.”

Religious types often think of the consciousness as the soul, but usually change their mind if forced to think critically about the subject. Then they realize that they consider the soul more mystical. Most of them would consider an extremely deteriorated person (like an advanced Alzheimer patient or a brain dead person) to have little or no consciousness, but would argue for the existence of a healthy soul.

Some seem to view the consciousness (and the mind) as mystical and perhaps not dependent on the physical brain. Believers in Astral projection, channeling, communication with the dead, and other such nonsense obviously consider both the mind and the consciousness capable of existing when spatially separated from the body.

It surprises me that we do not seem to have much meeting of the minds on what consciousness is. While it seems difficult to understand the mechanism behind it, I never thought the concept was that difficult to define. In spite of various posts to this thread, I think there a high level of agreement. There are some confusing semantic issues here, and some who will always try to muddy the waters of any already difficult subject.

As I posted earlier, the big guns who try to understand consciousness say things like the following.

The human mind seems to have utilitarian functions related to problem solving, survival, satisfying desires, et cetera. There also seem to be other activities of the mind which are much like the activities of the viewer (or critic) of a movie or the reader (or critic) of a book, where the movie being viewed or the book being read is the history of the individual. These other activities are called the consciousness of the individual.

When a human plays chess, there are thoughts like the following which are not directly related to playing the game

• I am playing chess, and I am bored with this game.
• I do this well.
• My opponent is a nice guy, but a lousy chess player.
• I should read a good book about Queen Pawn openings.
• I should get more exercise. I wonder if Joe can play tennis tomorrow.

When Deep Blue plays chess there are no such thoughts.

The consciousness does not seem to be required for the performance of any specific utilitarian function, but might be necessary for the mind to function at all.

I am convinced that my consciousness exists, and that all but extremely dysfunctional humans have a similar consciousness. I have no proof for these beliefs. I just know that I have it, and consider it absurd to believe that others do not also have it. Others seem to be similar to me in most of the characteristics I can observe, why should they be fundamentally different in the characteristics I cannot observe directly?

[Guv]

The Searle discussion was a tuff read.

It does not seem to me that Searle has proven that AI is impossible. I do think that he has presented a convincing argument that AI is not possible with computers using current architecture.

I think that AI is possible, but that it requires neural networks, or arrays of thousands or millions of CPU’s with some yet to be designed connectivity, or some other architecture unlike current systems.

It seems to me that Searle has not provided a formal proof, and I am not sure he was trying to develop a formal proof. Since the URL led to Larry Hauser’s article criticizing Searle, it is hard to figure out exactly what Searle said. Intent is tuff to determine under any circumstances. From what I read about the Chinese Room, I do not think it was an attempt at a formal proof, although the Hauser article indicates Searle formatted his argument like a formal proof, and used the Chinese Room in his argument.

It is fairly obvious that Hauser is attacking the Searle argument by showing that it is not a formal proof, which might not be fair if Searle never claimed to be presenting a formal proof. Hauser seems to me to be correct in claiming that the proof by Searle is invalid logic. However, Hauser does not seem to describe or refute the ideas put forth by the Chinese Room experiment.

For those not familiar with The Turing Test and the Chinese Room thought experiments, I will describe them.

Turing came up with the idea of testing future AI devices versus a human being. His idea is to put one or more people in a room with a keyboard and two monitors (actually in those days two crude typewriter like devices were proposed). Questions or remarks would be entered via the keyboard, and a pair of answers or pertinent remarks would be provided via the two monitors (teletype machines). One monitor would be controlled by a human and the other would be controlled by the AI device.

The people in control of the keyboard could ask questions or merely carry on a conversation with the entities in control of the monitors. If the people in control of the keyboard could easily tell which monitor was controlled by the AI device and which by the human, then the AI device failed the test. If the people could not figure out which was which, the AI device passed the test 100% and had to be considered intelligent. Obviously, it would be possible for the AI device to be given some score between 0 & 100%, depending on how difficult it was to identify which was which.

A full Turing Test would have no restrictions on the subject matter, and nobody expects an AI device to pass such a test with a high grade in the next 50-100 years. It seems reasonable to expect some AI device to pass a less rigorous test with restrictions on the subject matter or which otherwise gave the AI device a little help.

The Chinese Room thought experiment is an argument which claims that the Turing Test is not a valid way to assess an AI device. Imagine that you can pass Chinese ideograph text into a slot in the door of a room. Some time later, an English translation is provided via a second slot in the door. If a person fluent in Chinese and English verified that the translations were correct, it would seem reasonable to conclude that a person or an entity in the room understood both Chinese and English.

Now suppose that you investigated the contents of the room and discovered a group of people running computers. One group was looking ideographs up in a computerized database of ideograph images and creating a literal translation of original messages. Another group was processing the literal translations versus a huge subject indexed data base of Chinese ideograph literature with English translations. This group would use contexts found in Chinese literature to refine the literal translation. All the notes made by these two groups were passed to an output group which typed up the final translation.

After understanding the processes inside the room, most people would conclude that there were people in the room who understood English, but that nobody there understood Chinese. Even if the operations were carried out by one person, most people would not conclude that the person understood Chinese.

Note that the Chinese Room is similar in some ways to a Turing Test for comprehension of the Chinese language. The argument is that if you understood how a proposed AI device functioned, you might not credit it with intelligence. This is why Deep Blue is not considered an AI device even though it beat the best human chess player and chess playing is generally considered to require intelligence.

It seems that Searle carried this Chinese Room argument a bit further and claimed that AI was not possible.

[Kzin]

Some of you (and I'm thinking particularly of Guv & Parksie) might find this (from the AI labs at MIT) particularly interesting in this context

WHY PEOPLE THINK COMPUTERS CAN'T

[kedaman]

I hate to say this, but you look very ignorant to me, all of you. (except Kzin, I didn't distinguish his opinion)
guv

If you need to discuss or define what intelligence is, please start another Thread.

We still haven't defined what intelligence is, since a year ago when I started that thread. How can you relate to a undefined concept? ignorance yes...
simon

OK then, let's define awareness (as far as I know, no one's ruled out discussing that yet).

A computer gathers information through input devices such as a keyboard and a mouse. If your definition of awareness includes a computer by means of it's keyboard and mouse then I would say that your definition of awareness is different than mine.

i.e. A computer is not aware, even though it gathers information through various input devices.

I did go on to my specifically define consiousness/awareness as "the sensation of experience". This, I believe, more acurately describes my definition of consiousness.

And you imply that the computer is not feeling this sensation? If you ask it, it would answer you it had. Let's assume you have developed a computer to act as a specific human, it would say it feels this sensation, and the human would also say this. How do you distinguish between them?

Until you accept my theory, there's no way to proove. There's no way to think it's possible to proove.

I define consciousness as my own awareness, and a possible similar experience of others. We do apply consciousness concept on other human beings, and sometimes animals etc... but you can't proove whether a turing machine, computer, a dolphin, or even another human being possess consciousness, and we shouldn't put a line between them, they all can be either fully conscous or not at all.

Until then, this debate should be degraded to "which god is the right one" stage. That is, let your belief decide, if you wish. It's a religious matter.

Perhaps as far as you're concerned but it's hardly been resolved satisfactorally one way or the other yet. Still that's another side road I'm not going to go down now.

You know what I've told you Simon. It's the only fully explanatory theory that fit's both our consciousness and our enviroment, why hang onto something that is obviously incorrect?

[simonm]

Kedaman

And you imply that the computer is not feeling this sensation? If you ask it, it would answer you it had. Let's assume you have developed a computer to act as a specific human, it would say it feels this sensation, and the human would also say this. How do you distinguish between them?

I don't remember implying that a computer cannot have the sensation of consiousness. And no, I know of no way with which it is possible (externally) to determine whether something is consious or not.

Until you accept my theory, there's no way to proove. There's no way to think it's possible to proove.

Which theory is that agian? As far as I'm aware, this is the first post you've made in this thread so what are you refering to.

I define consciousness as my own awareness...

Fair enough. It's just that no one else has experience of your awareness so it is not a definition that is useful to everyone else.

You know what I've told you Simon. It's the only fully explanatory theory that fit's both our consciousness and our enviroment, why hang onto something that is obviously incorrect?

Now you've really lost me. What is it you've told me that has particular relevance here? What is it that I believe that is obviously incorrect?

[simonm]

I've just been reading the website that Kzin pointed out and I found the following very interesting statement:

So such concerns are not about
computers at all, but about our foolish quest for meanings that stand by
themselves, outside any context. Our questions about thinking machines
should really be questions about our own minds.

The first sentanceis a reference to our quest for absolute truth. It sums up my belief that truth is meaningless without context. i.e. Absolute truth may or may not exist, but it is only context which injects it with meaning.

The second sentance also echo's my belief in that when asking such questions about AI we need to first understand what makes us intelligent (or why we think we are intelligent).

But, I do think that generally, this guy grossly underestimates how difficult it would be to make a computer self-consious. Indeed, he states that it would be possible to make computers more self consious than we are!

I see no special problem in giving them the "self-insight"
they would need to understand, change, and improve themselves.

I think the above statement is based on a significant failure to understand what makes us self-consious.

I think this could be done by providing machines with ways to examine
their own mechanisms while they are working. In principle, at least, this
seem possible; we already have some simple Al programs that can
understand a little about how some simpler programs work.

But then what. Do you need ways to examine the mechanisms that examine the mechanisms? Where does this stop? Obviously, this could go on indefinitely and this is exactly what humans can do and what computers can't. We intuitively grasp an infinte and undefineable concept.

[csammis]

First off, I'm not a philosopher. I haven't the slightest idea of what consciousness consists of, nor do I know how one can prove consciousness, self-awareness or even sentience. Not going to pretend I do...I'm a computer scientist, I'm not paid to think about humans (j/k!)

Earlier in this thread someone made the point of mechanical vs. organic systems, and the quote about "unplugging eyeballs"...it's time that everyone here remembed what human beings are.

We are machines.

Completely elegant structures, remarkable creatures, capable of doing wonderful and beautiful things. But the fact remains, to a large extent we are processors of information, biologically based computational devices. Yes humans (most humans ) are capable of extropolating and "understanding" the data we receive, but we are still machines.

the computer is not intelligent attall. It just crunches numbers and follows a program.

I'd like to make my own amendments and continue:

the computer is not intelligent attall. It just crunches input and stored data and follows a set of rules.

In a very large-view and yet very real sense, that sounds a lot like a person looking at an art museum which has sadly caught fire. There are paintings, beautiful things inside which will be destroyed. There are still people trapped, but the firefighters will be there any second. The person's own safety is certainly something to consider to.

In this scenario, there are amazing parallels between the person and a computer as described in the quote. This person is considering the input (burning building, firefighters coming, people/treasures at risk) and possibly stored data (maybe he was burned as a child and is afraid of fire). He is running this input/data through a set pattern of rules, or morals ("Do I rush in and save people and art, risking myself, or do I wait in complacency while people may die?"). Think about how a computer processes input and stored data, and uses a program to tell it how to properly processes that data. Tell me how the person, in an amazingly complex situation, is not behaving like an advanced living breathing machine.

Are computers now any worthy comparison to the human brain? No. Binary data will never be able to match the fluidity of the brain for sheer flexibility. Can computers be built to break the Yes-No constrait and *really* acheive Yes-No-Maybe? I don't know. Binary structures and ways of thinking are drilled into our heads while learning about computers, and if you think about it everything a human does can be boiled down to "Yes" or "No". Take right now...I'm deciding whether to stop writing or not. My answer is "Yes, stop writing"

[simonm]

csammis

Tell me how the person, in an amazingly complex situation, is not behaving like an advanced living breathing machine.

One thing we have, that many people arrgue is un-computational, is intuition. It's what gives us common sense, our morality and our ability to grasp undefineable concepts such as numbers, infitity and the self.

How can we program intuition? I can't even begin to think about how I would program something so ambigous.

I think we need to persevere with studying what makes us consious. What gives us the sensation of being "awake". If we ever find a solution to this problem we just might be able to apply the right method to make a system truly aware of it's own existance and truly intelligent.

[kedaman]

Fair enough. It's just that no one else has experience of your awareness so it is not a definition that is useful to everyone else.

I didn't imply others couldn't have it, but if you want to refer to other consciousness, then it's your own consciousness that you extrapolate. Let's say a machine didn't have consciousness. It wouldn't feel this sensation of experience and therefore it can't relate to it, but if it works similar to one that has, what's the use declaring consciousness for it? If you are looking for moral aspects, it all comes down to primitive beliefs.

Which theory is that agian? As far as I'm aware, this is the first post you've made in this thread so what are you refering to.

Hmmm, we've been discussing this for some while (mainly in the thread "Benefits of science, religion, capitalism, & socialism" but also in some others) but you've probably had a lot else in your mind. My theory is that universe as it looks now doesn't exist, that it is a sub-reality, a dreamworld, a projection of our mind, and that if we accept the fact that it is, it will instantly disappear. For this to be any use in any discussion we can refer to different realities, which can have different sets of definitions and rules, but are constant and not contradicting within the same reality. This reality we think we live in we don't. That dismisses the consciousness aspects and everything in this world becomes physical, that's the only way we can relate to it too. In this discussion it would result in that there is no consciousness to talk about. The subjectivity of quantum mechanics can be applied under these circumstances, but we should be aware that we are aware (a bit pointless I know, but most people don't notice these doesn't match) As a result of this, we face contradictions, which are due to that this reality doesn't exist, but is a subreality of another. By accepting these contradiction (my way) you'll be able to distinguish what exists here, and what not.

The first sentanceis a reference to our quest for absolute truth. It sums up my belief that truth is meaningless without context. i.e. Absolute truth may or may not exist, but it is only context which injects it with meaning.

Look at what you say. The context to an absolute truth is out there, not here. What you have are just made up contexts, it has a meaning but it's at times incorrect. The absolute truth would provide you with absolutely true contexts.

possible to make computers more self consious than we are!

I'm quite sure he has no idea what he is dealing with, even if he think he knows.

But then what. Do you need ways to examine the mechanisms that examine the mechanisms? Where does this stop? Obviously, this could go on indefinitely and this is exactly what humans can do and what computers can't. We intuitively grasp an infinte and undefineable concept.

I think we are at this point no different than machines. We cannot examine ourself, only think we know what we are doing, only think we control ourself and make decisions, only have a context that is not even true.

[csammis]

I agree, this cannot be done with current 4th-generation computers. My point is that all the talk about how computers can be made more like humans is useless, because currently it simply cannot happen.

Anyone who knows quantum theory will understand this. Quantum computers are the answer, because human intuition has a lot to do with the principle of "quantum uncertainty." If a computer could solve the Schroedinger's Cat problem, that could be considered intuitive work. A normal computer wouldn't know the answer given the facts at hand. It's a certainty that the cat is dead, but a computer would be unsure until it had firsthand evidence. Humans (and as such a "5th-generation" computer) wouldn't know for sure, but intuition tells us that the cat is no more.

To sum up the case for AI: Binary = bad. Wetware (brains)/quantum computation = good

[Guv]

Kedaman: Do you post under the influence of something you injected, smoked, or snorted? I often wonder if you ever read what you typed before you post it.

My theory is that universe as it looks now doesn't exist, that it is a sub-reality, a dreamworld, a projection of our mind, and that if we accept the fact that it is, it will instantly disappear.

The above is not the first off the wall opinion I have seen you post, and I doubt if it will be the last. At least the above is understandable verbiage.

I often find it hard to take you seriously. Many of your posts seem almost unintelligible. Sometimes I am reminded of some of the Eastern Mystical crap I used to hear like the sound of one hand clapping. It merely sounds profound, but is actually nonsense. Most people are afraid to scoff at crap like that for fear of seeming stupid. It never bothered me to say that such drivel is meaningless because it is meaningless. At least Carroll did not pretend to be saying anything meaningful when he wrote the following.

Twas brillig, and the slithy toves
Did gyre and gimble in the wabe:
All mimsy were the borogoves,
And the mome raths outgrabe.

Next, it seems to me that almost every body posting here has a working definition of intelligence which is not far different from what others believe intelligence to be.

If you claim otherwise, you are really saying that any thread related to AI, consciousness, intelligence, the human mind, et cetera cannot be discussed in any meaningful fashion on this forum because we have no common understanding of the vocabulary required for the discussion.

I avoid forums which use Chinese, Swahili, et cetera. It seems to me that you should similarly avoid any threads discussing intelligence, since you claim that nobody has any knowledge of what intelligence is, or at least you claim that their view of it is radically different from your view.

Note that I did not assume that most of us have a similar understanding of the meaning of consciousness.

You constantly harp on the subject of provability. Everybody with any knowledge of formal logic knows that nothing is provable in an absolute sense. All so called proofs rely on axioms which are assumed rather than proven to be true. Without a proof of the axioms, none of the conclusions of any axiomatic system are absolutely provably true. You do not have to tell us this over and over again as though we are children who do not understand.

The Pythagorean theorem while almost universally accepted as true in Euclidean space, is not absolutely provably true. We do, however, have faith that our airplanes fly most of the time and that our bridges do not fall when we drive across them. For all practical purposes, the theorems of mathematics, physics, and science are valid now and will be valid in the future. Perhaps we will discover that all of our science cannot be extrapolated beyond the experimental evidence on which it is based. That is what happened to classical physics. It was not shown to be wrong. It was only shown to be not applicable to phenomena not known or not measurable at the time classical physics was developed.

In discussing subjects like AI, consciousness, the human mind, perception, et cetera nobody expects axiomatic proofs ala geometry and calculus theorems. At least I do not. The most we can hope for are arguments that are reasonable and which seem consistent with the meager applicable knowledge available to us. We can hope that though reasonable discussion we can improve our own concepts and opinions on difficult subject matter.

Being told that nothing is provable adds nothing to threads like this one. Claiming that any imaginable belief might turn out true in the future is a ridiculous copout. You harp on provability and use logically invalid arguments in your own posts. Time and again I notice you mentioning some stupid point of view held by somebody 500 years ago and use it to support a belief in the future possibility of some weird notion. Stupid opinions from the past do not support the credibility of any belief.

If you want precise definitions of all terms and absolute provability, stick to Euclidean geometry and you will get close to such an ideal state of knowledge. Note, however, that formal logic requires a set of undefined terms. Hence, you will never get every term defined. Stay away from discussions of AI, consciousness, et cetera or else present reasonable views backed up by some sensible commentary, instead of knocking the point of view of others because nothing is provable.

[simonm]

kedaman

I was going to reply to your post until I read Guv's post afterwards. You'll think twice before critisising his posts again?

I do remember our conversations in other threads, I just didn't think they were particularly relevant to this thread, nor did I say that I necessarilly agreed with them either.

Guv

The above is not the first off the wall opinion I have seen you post, and I doubt if it will be the last. At least the above is understandable verbiage.

While I'd admit that much of what Kedaman posts is difficult to understand (and usually has to be probed to express it more clearely), I think you're being a little harsh on him. English is still a relatively new language for him and these aren't the easiest subjects to talk about anyway.

Next, it seems to me that almost every body posting here has a working definition of intelligence which is not far different from what others believe intelligence to be.

If you claim otherwise, you are really saying that any thread related to AI, consciousness, intelligence, the human mind, et cetera cannot be discussed in any meaningful fashion on this forum because we have no common understanding of the vocabulary required for the discussion.

Well, since you were most adamant at the beginning that you didn't want this thread to degenerate into a debate about the definition of intelligence, you didn't leave us much choice but to agree on a working definition.

Because we didn't have a "common understanding of the volvabulary" there was a lot of (most likely) needless debate between people who were arguing from different paradigms (myself included).

csammis

To sum up the case for AI: Binary = bad. Wetware (brains)/quantum computation = good

I don't know enough about quantum computers to make a judgement on that but it most likely is a step in the right direction.

[Guv]

When I first thought about the relationship between consciousness and intelligence, it seemed to me that people did not differ much in their concepts of intelligence, but that consciousness was a far more difficult concept.

The big problem with intelligence seems to be measuring it and categorizing various components of it. Is creative ability more important than learning ability? Is an artist or composer more/less intelligent than a mathematician or a physicist? Are we born with it or do we acquire it? These seem to be issues that cause major disagreement, while differences over definition seem minor.

It has never occurred to me to wonder about other people being intelligent. Their behavior gives so many clues.

When it comes to perceptions and consciousness, I have no behavioral clues. Except for those who are color blind, we all agree to call this object red and that one blue, but I have no idea of what other people really see. If telepathy were possible and I could view the world through other eyes, perhaps the sky would look red and blood would look green.

Similarly, when it comes to consciousness or self awareness, I have no clues. Because I believe that I have it and other people seem similar to me in so many ways, I assume that they are conscious and self aware. I have nothing I consider behavioral clues to their consciousness, and a claim to have it is not convincing. If I were the only human in a world of incredibly capable robots, their behavior might convince me that they were intelligent. I cannot imagine how I would come to any conclusion about consciousness or self awareness.

Faced with an AI device that claimed to be self aware, is there any way I could verify or refute the claim? I could devise a lot of tests to verify or refute behavior which at least mimicked intelligence, but what test is there for self awareness?

To me, Deep Blue provides a hint of intelligence without consciousness. I do not see any direct connection between what I consider intelligent behavior and my consciousness.

I can imagine intelligence without self awareness, but not the converse. Because I can imagine it, is it possible? I think so, but my belief has little conviction.

While others disagree, I believe that consciousness or self awareness can exist in a non biological device. Therefore, it seems possible to me for an AI device to be have consciousness, but I have no idea of how to test for it.

[kedaman]

Guv
Some guys would agree, there was need for those like you as well as those like me for civilisation to advance.

Time and again I notice you mentioning some stupid point of view held by somebody 500 years ago and use it to support a belief in the future possibility of some weird notion. Stupid opinions from the past do not support the credibility of any belief.

If this in particular upsets you, you have missunderstood me. I was mentioning it a while ago, once, as an example, not as support for beliefs, rather to concretisize my theory. I was refering to evolution as whole (which was the actual support for my theory), do you deny it happened/is happening/is going to continue? I suppose you don't, so if you want to continue the discussion, we could, but not on this thread.

You do not have to tell us this over and over again as though we are children who do not understand.

I know, most people will never learn, how many times you ever tell them. But I remain hopefull. I agree with you that it is a problem to bring it up in every single matter in life, but sometimes, especially on this subject, I think that it will bring some sense, instead of nonsense. If there aren't any other reasonable theories, why not accept the possibility of the radical?

I often find it hard to take you seriously. Many of your posts seem almost unintelligible

I didn't expect you or anyone in this forum to take me seriously when I wrote that specifically. I was explaining it to simon earlier and it took me several pages for him to even accept it is understandable. The last one I expect to try to understand me is you, so I didn't consider doing any explanation here. If you've found related posts by me you might have encountered the same theory. Since I expect a serious response, I'm adapting this post specifically for your convenience. Just ignore my theory for now.

Next, it seems to me that almost every body posting here has a working definition of intelligence which is not far different from what others believe intelligence to be.

Imagine a group of knowledgeable men working on trying to proove certain paragraphs in the bible. They would not accept a scientist coming around and say that Earth is round or something, he'd be hanged immediately. I'm not a scientist around here, but what's wrong being openminded?

I avoid forums which use Chinese, Swahili, et cetera. It seems to me that you should similarly avoid any threads discussing intelligence, since you claim that nobody has any knowledge of what intelligence is, or at least you claim that their view of it is radically different from your view.

If an enough intelligent being would take your advice (for some odd reason) he would go and hang himself because he don't want to participate in life. I'm not going to hang myself, first because I'm the optimistic guy, and second since I feel participating might trim my intelligence further.

In discussing subjects like AI, consciousness, the human mind, perception, et cetera nobody expects axiomatic proofs ala geometry and calculus theorems. At least I do not. The most we can hope for are arguments that are reasonable and which seem consistent with the meager applicable knowledge available to us. We can hope that though reasonable discussion we can improve our own concepts and opinions on difficult subject matter.

To me it sounds like you are contradicting yourself. Trough "Reasonable" discussion we apply logic and theorems all the time, if it's not directly mathematical, it still is built up on logic, otherways we would not be able to discuss anything coherent.

Anyway, I apologise for trying to knock off the definitions you've set up for this thread. Instead of relating to my vision, you might consider this reasonable:

Moral aspects vs Scientific aspects
IF there's a purpose for AI, you probably agree with me that it's not we that are to be replaced by it, especially you. Say tasks currently operated by other human beings can be operated by AI devices. If that's what you need AI for, then there's no need to drag in consciousness in the business, if you are the scientist. If you're the moral guy, you have to concider about consciousness though, and ask if AI obtains consciousness, and if so, should it be given human rights and so on... Now which discussion are we leading here? The scientific right?

The scientist would look at it this way: Human beings are no different from machines, they are just biological machines. Therefore it's possible to replicate machines that emulate humans, or at least perform tasks that can be performed by humans.
The moral guy would say: Human beings have consciousness and therefore machines cannot be replicated to emulate humans, which probably results in not being able to perform the tasks either.
The philosopher might make an observation here: While there's no proof of other human beings to have consciousness, machines not induce consciousness on their own, consciousness should not be assumed to be needed when implementing the AI device.

Concluding that, consciousness should not be taken as relevant if you intend to develope AI. If our intentions are to provide the same results as intelligence can provide, an emulation that does the same will do. Since you asked for everyones opinion on the relation between intelligence and consciousness, I think you have to accept responses like this. If you would have asked how consciousness relates to the developement of AI, my answer would have been straight forward and the missunderstandments would never had taken place.

If csammis is pointing to the right tools we need, then it's those we need to discuss.

[Nucleus]

Perhaps I am oversimplifying things too much, but a prerequisite for AI is we as humans to completely understand how our cognitive processes work.

This has yet to be achieved. Perhaps such understanding would explain consciousness?

Assuming we will be successful, we then have to work out how to replicate these cognitive processes. Like csammis, I don't think binary computers will be the vehicle to achieve this task.

Once this is achieved (how easy that was to say ), to me the exciting thing is that we can be proactive in improving the design. Although this requires us to overcome our own ego as it means we would be creating a "computer" (probably not what they will be referred to as) with better cognitive processing than our own. I hope we are able to achieve AI and beyond.

[csammis]

Perhaps I am oversimplifying things too much, but a prerequisite for AI is we as humans to completely understand how our cognitive processes work.

Contrairlywise, if we stumble across the secret of AI it would vastly increase our understanding of ourselves. I still believe that by building iterative models we can understand the larger part...since I've made my point that brains are truly advanced computers, perhaps building bigger and better computers is the way to find out about the way we work.

As to bigger and better computers...my computer weighs more than I do, and has 10.3 cubic feet of storage volume. Does that mean that my comp. is heading towards understanding my brain? Nah

[simonm]

Guv

Faced with an AI device that claimed to be self aware, is there any way I could verify or refute the claim? I could devise a lot of tests to verify or refute behavior which at least mimicked intelligence, but what test is there for self awareness?

I think the right direction for such tests would be along the lines of Godel's incompleteness theorems. If you asked the device what it would do in a new situation (that it is not curently familiar with) then it could only answer if it had a concept of self. If you had to feed in a complete description of itself (in order to answer the question) then it would not be self-consious and would not be able to answer the question. i.e. It would have to run a sub program that was a simulation of itself and that sub program would in turn have to run a sub-sub program etc. to infinity.
Only an entity that had an intuitive grasp of itself could ever come up with an answer in a finite amount of time.

Obviously, the quesiton would have to be picked very carefully to avoid the device's creators being able to trick their way out of such questions.

I can imagine intelligence without self awareness, but not the converse. Because I can imagine it, is it possible? I think so, but my belief has little conviction.

Do you consider chimpanzees intelligent? They have been shown to be self-aware although they are (as far as we know) much less intelligent than humans.

All in all, I think there are limits to how far you can go with intelligence without self-consiousness. Without the ability to think about oneself in the third person, then it is simply impossible to postulate about oneself in new situations.