Is consciousness necessary for intelligence?

[Nucleus]

Assuming we as humans are both conscious and intelligent, is consciousness part of our cognitive process hardwired when we are born (are we born both intelligent and conscious), or is consciousness something that may be learned?

I don't think many would argue we as humans are born without intelligence, for me this is the ability to learn.

So are we are born conscious or not?

If we are born conscious then this potentially supports a codependency between intelligence and consciousness. If not then this supports intelligence not requiring consciousness.

I gravitate towards thinking that intelligence makes consciousness possible, rather than consciousness being required for intelligence.

[simonm]

Nucleus

I don't think many would argue we as humans are born without intelligence, for me this is the ability to learn, although I could be mistaken.

If intelligence is the ability to learn then we at our most intelligent when we are newborn and at our most stupid when we are old.

I heard a statistic once that said we have learned 90% of what we will learn (our entire life) by the time we are 5 years old.

So are we are born conscious or not?

I think we are not only born consious but we have it before we are born. At exactly what stage of phoetal development this happens I don't know but there is evidence that babies remember music they here in the womb (for example).

As for when we develop self-consiousness, that's probably a lot more difficult to determine.

I gravitate towards thinking that intelligence makes consciousness possible, rather than consciousness being required for intelligence.

This probably depends on your definition of intelligence, and as far as I can make out, your definition differs from mine (I don't really have one but I know it's not the same as yours).

[Nucleus]

simonm,

Yes by consciousness I mean self awareness as everyone seems to be discussing this concept.

Consciouness as you describe it is the ability to take in information through our senses?

Yes we are born conscious in the sense that we hear, feel hungry, see, smell, and importantly think as soon as we are born if not before as you suggest.

In this sense of the word I agree intelligence and consciousness are linked.

However, at what point do we become self aware?

[simonm]

However, at what point do we become self aware?

I think we're comming back to the general question of how to determine whether anything is self aware.

I think the test we apply to animals is whether they are able to recognise their own reflection (as being themself). Apparently, Chimanzees and Dolphins have given indications that they are self-aware but as for how reliable this test is, I do not not know.

How and why something moves from being merely aware to being self-aware is a mystery in itself.

Yes by consciousness I mean self awareness as everyone seems to be discussing this concept.

Consciouness as you describe it is the ability to take in information through our senses?

I have read many books and scientific papers on the subject and they (in my experience) tend to make a distinction between being having consiousness and self-consiousness.

[Nucleus]

Simon, I just re-read your first few posts on this thread, Guv defined consciousness as self-awareness, you preferred to distinguish awareness from self-awareness.

Ok.

If we are not born self-aware, then perhaps this would support the case for self-awareness not being required for intelligence?

[simonm]

If we are not born self-aware, then perhaps this would support the case for self-awareness not being required for intelligence?

Perhaps but how do you tell if a baby is self-aware (especially before it is born).

Until we discover exactly what form of brain development is required for self-consiousness to manifest itself, we may never be able to tell if an unborn baby has it or not.

[Nucleus]

Simon

I guess I found it difficult to grasp the concept of a newly born child being considered self-aware, and consequently gravitated towards a belief that self-awareness was not required for intelligence.

You are right there is no way at the present to confirm this to be true, so it can only be considered speculation.

[csammis]

What is being argued right now is the exact reason why there is so much controversy over abortion. Is a fetus self-aware? Cells aren't self-aware, and yes this is an oversimplification but a fetus is a batch of replicated cells. At what point can a sac of cells rise up and become conscious or self-aware, or can there even be a set point? Does a baby wake up one day and say "Cognito, ergo sum", or is it more of an evolutionary processes in which experience is gained through sensory apparatus, pathways and connections are formed in the brain and eventually the baby becomes self-aware? Studies show that a baby is very much aware of its evironment after being born, but as to whether it is self-aware we may never know. One would have to be a mind reader to answer that, since a baby can't exactly communicate what's on its mind

[TKarr]

Cells themselves obviously are not self-aware or we might have a war inside our bodies, in addition to the current daily fight of disease and harmful bacteria vs healthy cells. The collection of cells creates a being, which due to the right combination and interaction of luck, dna, and cell matter has self-awareness.

This conversation could last forever. Here's my tangent:

How many of you people have seen comatose patients, TV or not? Yeah that's what I thought. Now...did that person know he or she was alive? We don't know, unless they wake up. Some people don't even remember that they were in a coma. BUT, we, as non-comatose people, have two choices. 1) Say "Yes, that person is alive." 2) Say no, he or she is braindead, and therefore not alive. (And technically there is a third: "I don't know/care")

Back to the...dare I say it?...abortion issue!

Now the "fetus" as people like to call "it" starts out as a few cells, yes. But we too are a few cells, multiplied millions upon millions of times. This collection of cells forms an unborn baby. Wow aren't we all geniuses? I bet you didn't know that!

Now imagine the baby after it has formed a head, torso, arms, and legs. Look at yourself. See how you and the unborn baby both have those parts? Ta-da. Soon-to-be-born human. Now imagine yourself when you were that tiny. Hard to do? Try imagining intruments ripping you away from your life source-- your mother. Of course, now half of you (at least) are saying, "but I wouldn't know, because I didn't have self-awareness."

Have you ever looked at a live ultrasound? The baby kicks-- moves. The baby touches himself or herself. Yes, a machine can do the same. Yes, we can program it to do so. But like was said before, humans are advanced machines. They just get more advanced--evolve, if you will--as they get older. If the baby were to touch the wall surrounding him or her, would it be out of reflex or would it be because he or she was saying, where am I? Surely the brain at that point would not be able to remember anything at that point. It only has a few cells. The baby probably has to re-write over his or her memory at that point.

::Move hand:: = stored for seconds, gradually building time to eventually add to motor-movement area of the brain
::Touch surroundings:: = stored for a few seconds, to prevent repetitive action, eventually building to eventually become the "I need to get out" feeling the baby has before birth.

And so on and so on. Multiple occurances happen at once, which adds to the possiblity that the baby is self-aware.

YOU were once a tiny few-celled organism, hard as it may be to believe. And look at you now. Self-aware as ever! What may not be self-aware before birth, becomes self-aware eventually.

One of the things a person must have to be intelligent is the ability to make decisions. The unborn baby has to make the decision to move or touch, or else it's just reflex. So if the baby knows he or she exists, he/she can move his/her arms and legs because he/she knows they're there and can make decisions with them (i.e. touching surroundings).

self-awareness ----> decisions----> intelligence

[HarryW]

You seem to be arguing your point backwards, self awareness requires intelligence. Any kind of awareness requires at least some level of intelligence. Intelligence isn't the issue here, fruit flies are intelligent but nobody campaigns outside shops that sell fly spray, nobody beats up the shopkeepers for selling it.

There are, quite obviously, varying degrees of intelligence. I can't say whether a foetus has any level of intelligence above that of a fruit fly, because I have no idea. I am pretty sure that whatever intelligence they have is minimal though, and my hunch is that they are far below the level where they become self-aware.

I really don't think your extrapolation back to your own life in the womb is valid. The similarities between you now and you as a foetus are really no greater than the similarities between you now and any animal pre-birth, except you're the same species as the human foetus.

What's the latest that an abortion can take place legally anyway? I don't even know if it's legal anywhere in the US. I doubt it's legal anywhere like Texas, they seem wildly sensationalist and excitable around there.

[barrk]

Abortion is still legal in the United States and it has been upheld by the Supreme Court in test after test.....prior to four months pregnant. Our glorious leader is trying to do everything he can to change this but has at yet only managed to stop federal funding for clinics outside the contiguous states (air bases, etc. in other countries that are still considered US soil).

[kedaman]

Self Awareness vs Awareness vs Conscious.
It's quite hard to work without definitions.

If something is aware it can take input.

If something is selfaware it can take input for it's own conditions. A bit of an extension for aware.

If soemthing is conscious, it feels the sensation of experience.

The third is very unlike the first two, since it can't be directly determined, and there's no reasonable knowledge on the subject. The two other can be thought as independent of it or as a entity. Therefore it shouldn't be discussed in a matter like this either.

If something is intelligent, it can process information.

Now with those definitions can you create a machine that validates all awareness, self-awareness and intelligence? I don't see any problem. However human brain is a massive complex thing, with restructuring and evolving features, would it be a good idea to make a deep copy, or would it be better to replicate the functionality of it?
AI usage
Are we really ready to accept the risks of getting outclassified? Are we ready to dump the hard work of billion years of evolution? Wouldn't it be better to try to get off our lazy asses and think ourself? You might consider biological evolution slow, would a
cybenetic variant outevolve us? In that case I think there's enough moral in us to not let that happen. Evolution accommodated moral and ethical rules since individuals in societies are better off to survive and to keep societies coherent, moral laws are To not let ourself get extinct there are moral laws for judging whether AI is such a good idea.
Being lazy, tired of life?
Since we're an evolving species, we have the ability to get rid of our bad feautures and implement new ones, instead of pushing over the burden of "thinking" on machines, shouldn't we enhance that feature of us? Otherways we might get dumber than our ape-ancestors. This might just be the peak of our intelligence evolution. If computers are so good at calculations then why not commit fusion?

[csammis]

HarryW:

I really don't think your extrapolation back to your own life in the womb is valid. The similarities between you now and you as a foetus are really no greater than the similarities between you now and any animal pre-birth, except you're the same species as the human foetus.

I may be going out on a small limb here, but you and I are the same species. And yet, we are not alike. However, I am in fact the exact same person as the fetus I once was. More associational pathways have been forged yes, but Fetus Chris and Now Chris are more alike than you and I or any other human could ever be. If you are not like yourself, who are you like?

Intelligence isn't the issue here, fruit flies are intelligent but nobody campaigns outside shops that sell fly spray, nobody beats up the shopkeepers for selling it.

You consider fruit flies intelligent? What do you consider stupid?? Fruit flies display nothing but instinctual behavior, and that is hardly conscious/intelligent/self-aware/anything. And people do in fact campaign for animal rights, these same people probably don't spray their fruit flies.

TKarr:

self-awareness ----> decisions----> intelligence

Your progression may very well be what we may use one day to classify a true AI computer. Nobody can classify today's computers as intelligent, although they follow a pattern of binary "decisions". But, if someday a system is built that we would consider self-aware (however that may happen!), it would potentially display the ability to make complex decisions (see my earlier post in this thread re: man and burning building) and that could be considered intelligent behavior. HarryW, you are considering the humanistic view, and very few people would argue that humans do not have the capacity for self-awareness/intelligence/consciousness. TKarr has followed with my own mechanical view (that is, proving intelligence/consciousness in machines) and reminded us that the only thing that humans consider truly without any life are the computers, and thus should be a potent point of debate on this thread. I don't say this only because I'm a computer scientist...although I would like to know what HarryW and TKarr do for a living, this would allow for an interesting perspective!

[Nucleus]

Originally posted by kedaman
[B]Self Awareness vs Awareness vs Conscious.

If something is intelligent, it can process information.

Then by this definition computers are already intelligent. I would say a key element of intelligence is the ability to adapt and learn and to think in different ways.

At present perhaps computers can be programmed to learn a set topic, but they have yet to be programmed with the ability to learn in a generic sense.

If you can learn in a generic sense then it follows it is possible to learn about one's self in relation to the universe i.e it is possible to learn self awareness.

For me it is intelligence and in particular the ability to learn in a generic sense that permits the possibility of self awareness.

AI usage
Are we really ready to accept the risks of getting outclassified? Are we ready to dump the hard work of billion years of evolution? Wouldn't it be better to try to get off our lazy asses and think ourself? You might consider biological evolution slow, would a
cybenetic variant outevolve us? In that case I think there's enough moral in us to not let that happen. Evolution accommodated moral and ethical rules since individuals in societies are better off to survive and to keep societies coherent, moral laws are To not let ourself get extinct there are moral laws for judging whether AI is such a good idea.

Just because computers may evolve to be more intelligent and potentially more self aware than us why to we have to prevent this from happening?

I guess we are the most intelligence form of life on earth and look how we treat forms of life less than ourselves. Perhaps one can extrapolate that a more evolved form of life would seek to dominate the world and us in a similar way? If this be the case then we should ensure we never delelop AI to the point of being able to learn.

I am attracted to the concept of a more evolved form of life in an abstract sense, but I guess I have to temper this attraction with the potental negative impact it would have on our own species. Here we are both operating under the assumption that a more evolved form of life will attempt to dominate all forms of life beneath itself.

As an extension to this assumption, should we consequently avoid developing AI to the point of being able to learn, lest it out evolve us almost immediately? A further corollary would be that we should avoid trying to contact life elsewhere in the universe lest we encounter a more evolved civilisation?

Being lazy, tired of life?
Since we're an evolving species, we have the ability to get rid of our bad feautures and implement new ones, instead of pushing over the burden of "thinking" on machines, shouldn't we enhance that feature of us? Otherways we might get dumber than our ape-ancestors. This might just be the peak of our intelligence evolution. If computers are so good at calculations then why not commit fusion?

If you accept the assumption stated above then this follows nicely. This may come from a fusion of computers and humans, or it may come from genetic engineering or a combination of the two.

[simonm]

Nucleus

If something is intelligent, it can process information.

Then by this definition computers are already intelligent. I would say a key element of intelligence is the ability to adapt and learn and to think in different ways.

As we seem to be talking about defining intelligence now then I will give my offering.

I would suggest that perhaps Kedaman's definition is to basic and that Nucleus' definition is too advanced.

How about: Something is intelligent if it can turn raw data into information. i.e. it can extract meaning from chaos.

This would mean that computers (as they currently stand) are not intelligent but it would also classify animals (who are not self-aware) as intelligent.

And for any out there who would argue that computers do turn data into information, I would say: No they don't. They are tools for allowing us to turn data into information.

[HarryW]

csammis:

Fruit flies are intelligent. I'm not talking really intelligent here, but they are intelligent. They perceive the world around them, make decisions in what limited capacity for cognition that they have, and act on them through the use of their bodies. That is intelligence, at it's most basic level. Their actions are of course instinctive, since they have very little memory so they can't base their decisions on any more knowledge than what they were born (hatched? pupated?) with.

Note that I'm not saying they are self-aware, or conscious, but from my AI-related academic studies, I would consider them fundamentally intelligent. 'Stupid' just means 'not very intelligent'. It's not an opposite.

I'd be interested to know if you've ever come across or heard/read about anyone campaigning for rights for fruit flies, or anything similar that isn't a rare species. Even objecting to their use in scientific experiments.

I may be going out on a small limb here, but you and I are the same species. And yet, we are not alike. However, I am in fact the exact same person as the fetus I once was. More associational pathways have been forged yes, but Fetus Chris and Now Chris are more alike than you and I or any other human could ever be. If you are not like yourself, who are you like?

I think you misunderstand what I mean by alike. I mean physically and mentally alike, not genetically alike. I am almost sure you would not say that a foetus version of you had more similar patterns of thought or levels of self-awareness than you and I.

I don't mean to sound cold about it, but I don't see any big difference between a human foetus in early stages of development, and any other quadroped foetus in the early stages.

Nucleus:

If something is intelligent, it can process information

Then by this definition computers are already intelligent. I would say a key element of intelligence is the ability to adapt and learn and to think in different ways.

That's not what Kedaman said. His sentence can be roughly traslated into propositional logic las follows:

intelligence => can_process_info

but this alone does not imply the opposite, which would be
can_process_info => intelligent

it only implies this:
¬can_process_info => ¬intelligent

This is a one-way implication, if something is intelligent then it must be able to process information. It does not mean that if something can process information then it is intelligent.

[kedaman]

Nucleus

Then by this definition computers are already intelligent. I would say a key element of intelligence is the ability to adapt and learn and to think in different ways.

A computer doesn't process information on it's own, software process information using available hardware. Software are intelligent to a certain level, depending on the programmer. Take for instance the bots in Unreal tournament, they don't exactly stand still and get killed, especially if you play God-like.

The ability to learn doesn't imply intelligence, neither intelligence does imply the ability to learn, Learning means the ability to store information for later usage.

For me it is intelligence and in particular the ability to learn in a generic sense that permits the possibility of self awareness.

Think about what you say. self awareness is the ability to recieve input of it's own state. That shouldn't require anything but a hardware device.

Just because computers may evolve to be more intelligent and potentially more self aware than us why to we have to prevent this from happening?

We are social evolutionary species, and have morals, did you read what i wrote?

If this be the case then we should ensure we never delelop AI to the point of being able to learn.

What's the use of AI that can't store information?

Simon

How about : Something is intelligent if it can turn raw data into information. i.e. it can extract meaning from chaos.

information is interpreted raw data, the raw data is just a way to store information, and has nothing to do with intelligence. Processing is always operated on information, whether they are in form of neural impulses or bits.

And for any out there who would argue that computers do turn data into information, I would say: No they don't. They are tools for allowing us to turn data into information.

Software process information, and this information is interpreted from raw data. Information is a bit hard qualified subject, it's how you relate to raw-data. It's like the observer concept, there is no information without an observer. Software do relate to raw data for our convenience. A POV perspective can be assumed without implication of any consciousness, how can we otherways live in a society?

intelligence => can_process_info

but this alone does not imply the opposite, which would be
can_process_info => intelligent

it only implies this:
¬can_process_info => ¬intelligent

I have to correct myself. This is what I meant:

can_process_info == intelligent

Intelligence is not a fusion of two abilities or more, it's the very essential part of information in POV perspective, what use it is of defining information. Answer: It can be processed, so to speak applied on information, giving further information for the perspective owners convenience.

If there's anybody not agreeing with me now, please state your versions of definition and I'll explain eventual flaws.

Now on for Intelligence degree. It's also a concept depending on perspective. Information depends on perspective ^ Intelligence is the processing of Information => Intelligence depends on perspective. (note this what I just stated is an example of intelligence, how to process two pieces of information into a third.) Since Intelligence level depends of the perspective, anything that process information the way you do is intelligent, the less similar it is, the less intelligent it is. If someone stated that 1+1=3 then you surely must think he is unintelligent, that just happens to be flawed logic, however logic is a set of definitions you adopted. To him who think 1+1=3 is correct just because he has adopted it as a pure axiom, you seem unintelligent to state 1+1=2.
The complexity of rules you have adopted or the "flawlessness", "beauty" or "symetry" has nothing to do with intelligence degree. You decide yourself how much intelligence you need but in the long run it will be adapted to process the information based on your environment, since that would be most usefull.

[simonm]

Kedaman, I think we're getting at the same thing (but from slightly different angles).

information is interpreted raw data, the raw data is just a way to store information, and has nothing to do with intelligence. Processing is always operated on information, whether they are in form of neural impulses or bits.

I didn't say information was intelligence. I said that a device should be considered intelligent if it can turn raw data into information.
Raw data is objetive and meaningless. Information is subjective and meaningful in a particular context.

Software process information, and this information is interpreted from raw data.

Software may process information (if that is the task that it is programmed to do) but it does not interpret information. The interpretation is provided by the programmer and the software is a mere tool for the device.

e.g. When I hammer nails into a piece of wood, who is banging the nail? Although it is the hammer that actually bangs in the nail, it is me that is actually doing it. The hammer is just a tool to make my efforts more effective.

It's like the observer concept, there is no information without an observer.

I would qualify this further. It is not information without an intelligent observer. Unless the observer can interpret the data, it remains data.

[kedaman]

simon

Kedaman, I think we're getting at the same thing (but from slightly different angles).

I didn't say information was intelligence. I said that a device should be considered intelligent if it can turn raw data into information.
Raw data is objetive and meaningless. Information is subjective and meaningful in a particular context.

Maybe we just are. If you think information is meaningless without raw data then we are, otherways we have endless information but no raw data. No objectiveness, only subjectivity. Let's just say we talk information here, "device independent information" so to speak.

Software may process information (if that is the task that it is programmed to do) but it does not interpret information. The interpretation is provided by the programmer and the software is a mere tool for the device.

Applied Scope
Let's assume perspective is a concept that can exist without consciousness. You can point at any point in space and say it has a "vision" that it can see different things depending on where it is, even if is no "eye" placed there. This is what you do when you talk to other people, you assume that the other man/woman you are talking to have a perspective and that it is intelligent and react to what you say and so on... You do that without any provided proof that this man/woman actually is conscious, is not only a program or an illusion of your mind. You just assume it is so, because it is intelligent (observ my definition on intelligence degree).

Now if you assume the "tool pespective" everything in your enviroment turns into tools, you become a solipsist. The information is always only interpreted as your information, you friend is not providing information to you, he just act as a tool to retrieve information. He does not process information, he process raw data.

This is not applied scope, this is not how a social being should interpret it's environment, therefore we should leave this and stay consistent. If you have a solipsistic view, there's no need for anything except tools anyway.

I would qualify this further. It is not information without an intelligent observer. Unless the observer can interpret the data, it remains data.

mixed up definitions.
if there's no data we have only information, if we have data and information we are observers, what is for sure is that we have information, and so is that we have intelligence (by definition). That implies there is no dependency of observations. Your definition is a further qualification but it does not work under all circumstances, I would consider a machine concluding 1+1=2 from mathematical axioms intelligent, the axioms being pure information.

Merging concepts won't help you, especially when the first is objective and the second is subjective. If you want to argue on that, we could do so all night, I have plenty of time.

[simonm]

Merging concepts won't help you, especially when the first is objective and the second is subjective. If you want to argue on that, we could do so all night, I have plenty of time.

Well, I'm not entirely sure what you mean by me "merging concepts" but I will talk about it further. I get the feeling I don't have as much time to talk about it as you do but what the hell.

If you think information is meaningless without raw data then we are, otherways we have endless information but no raw data. No objectiveness, only subjectivity. Let's just say we talk information here, "device independent information" so to speak.

I don't think information is meaningless without raw data. Rather, information is interpreted raw data. I'm not sure if I think that information can exist without raw data...but I doubt it. If there is no raw data, there is only interpreted nothingness and that would imply that only I exist and everything else I see and hear around me are illusions created by myself for my own entertainment.

if there's no data we have only information, if we have data and information we are observers, what is for sure is that we have information, and so is that we have intelligence (by definition).

OK, I was assuming that information can only exist given both raw data and the observer. I assumed that the universe is not just an illusion constructed by my own imagination. I doubt if it is ever possible to prove that anything (outside of our perception) actually exists but there is no point in continuing this thread unless we assume (for the time being) that there does exist such a thing as raw data.

[Nucleus]

Simon you say

And for any out there who would argue that computers do turn data into information, I would say: No they don't. They are tools for allowing us to turn data into information.

Can you expand on this please I don't quite grasp your point. I think I need clarification on what constitutes turning data into information and why computers are incapable of doing this.



Harry,

This is a one-way implication, if something is intelligent then it must be able to process information. It does not mean that if something can process information then it is intelligent.

Ok I like the logic you give here. Ok, well what allows computers to process information?

Also what does ¬ literally translate as?


Kedaman,

A computer doesn't process information on it's own, software process information using available hardware. Software are intelligent to a certain level, depending on the programmer.

For a start I consider that software and hardware together form a computer. Hardware alone does not constitute a computer, nor does software in isolation.

Learning means the ability to store information for later usage.

Learning is not simply to have a memory, how can one learn anything if one merely remembers what happened? I touch a boiling kettle, it burns my skin. Without the ability to learn I cannot change the way I interact with the environment even though I can recall the interactions. With memory alone I would never learn not to touch a boiling kettle. Memory is however a prerequisite for learning.

Think about what you say. self awareness is the ability to recieve input of it's own state. That shouldn't require anything but a hardware device.

A hardware device can create self awareness . For me self-awareness is more likely software than hardware in so far as it is neither an input device nor an output device but rather part of the cognitive process.

quote:
--------------------------------------------------------------------------------

Just because computers may evolve to be more intelligent and potentially more self aware than us why to we have to prevent this from happening?

--------------------------------------------------------------------------------


We are social evolutionary species, and have morals, did you read what i wrote?

Yes I did read what you wrote, did you read the paragraph after this question where I go on to say that for our own self preservation we should prevent this from occurring based on the assumption that a more evolved form of life will attempt to dominate all forms of life beneath itself.

Did you read this? If you did you should have realised that I asked the question in order to provide my thoughts on this question, not because I disagreed with your POV. In fact I think I supplied thought from another perspective that augmented your POV.

What's the use of AI that can't store information?

Now you have really lost me, where did I say this My point was not to create AI with the ability to learn and thus out evolve us. Again it is not the ability to recall past events that by itself will cause computers to out evolve us, but their abililty to learn. So I would certainly not advocate AI that isn't allowed to store information, but AI that is not allowed to learn in a generic sense of the word.

[simonm]

Nucleus

Simon you say

And for any out there who would argue that computers do turn data into information, I would say: No they don't. They are tools for allowing us to turn data into information.

Can you expand on this please I don't quite grasp your point. I think I need clarification on what constitutes turning data into information and why computers are incapable of doing this.

OK, what I'm going on about here is to do with information theory. According to information theory, a statement contains one bit of information for every yes/no quesiton it answers. If you already know the answer, it contains no information. Therefore, I am equating information with Interpreted Data. e.g. A statement may contain information to me but not to you (as you may already know about it) so it is subjective.

Now the process of turning raw data into information is a process inside the mind of the observer. The raw data doesn't really change and the information doesn't really exist (in an objective sense). This process involves interpreting raw data by extracting meaning from it.

We are intelligent because your brain turns raw data (comming in through the senses) into information by a process of interpretation. The fact that we can extract meaning from raw data (even if it might be an incorrect interpretation) makes us intelligent.

[kedaman]

Simon

I don't think information is meaningless without raw data. Rather, information is interpreted raw data. I'm not sure if I think that information can exist without raw data...but I doubt it. If there is no raw data, there is only interpreted nothingness and that would imply that only I exist and everything else I see and hear around me are illusions created by myself for my own entertainment.

Raw data - If there is an objective universe, we assume it is the raw-data, but since we don't know, the device interface which supplies our information could be defined to work for any environment. If we just forget about raw-data and concentrate on information, we just have to adopt the applied scope perspective.

but there is no point in continuing this thread unless we assume (for the time being) that there does exist such a thing as raw data.

In fact I think, we just need to adopt applied scope, and everything will be just fine. Even if we can't produce something "real" we do produce what we want.

We are intelligent because your brain turns raw data (comming in through the senses) into information by a process of interpretation. The fact that we can extract meaning from raw data (even if it might be an incorrect interpretation) makes us intelligent.

the essense of information
You might consider raw data as information itself, just unprocessed information. If that's how we see at it, we are in agreement, I think. Both our views will give the same definition of intelligence, although you assume interpretation is part of it while I assume it's part of the processing as well.



nucleus

For a start I consider that software and hardware together form a computer. Hardware alone does not constitute a computer, nor does software in isolation.

Assuming this, computers are by definition intelligent.

Learning is not simply to have a memory, how can one learn anything if one merely remembers what happened? I touch a boiling kettle, it burns my skin. Without the ability to learn I cannot change the way I interact with the environment even though I can recall the interactions. With memory alone I would never learn not to touch a boiling kettle. Memory is however a prerequisite for learning.

Mixed up definitions here. When you conclude that it is a bad idea to not to touch the kettle, you process information, that is use intelligence. If you have the ability to learn, that is store this information you processed, you can use this information later on, to process again, the information that you see a kettle, and recall that you shouldn't touch it.

A hardware device can create self awareness . For me self-awareness is more likely software than hardware in so far as it is neither an input device nor an output device but rather part of the cognitive process.

I don't quite understand what you want to say. Self awareness, to me means to be aware of itself, there is an obvious implication of an input device.

did you read the paragraph after this question where I go on to say that for our own self preservation we should prevent this from occurring based on the assumption that a more evolved form of life will attempt to dominate all forms of life beneath itself.

Yes, but I don't quite see the relevance.

I don't think you disagreed with me but I thought you did not see my point, well nevermind, why should this be relevant?

So I would certainly not advocate AI that isn't allowed to store information, but AI that is not allowed to learn in a generic sense of the word.

generic sense of the word? I know what you want to say even if it's the wrong terminology you use. You mean that AI would start making conclusions on what it percepts and then use these as axioms in a war against us. It's not "learning" that is responsible here, it's the processing part, the intelligence of AI that might overthrow us.

[Nucleus]

Simon,

Software may process information (if that is the task that it is programmed to do) but it does not interpret information. The interpretation is provided by the programmer and the software is a mere tool for the device.

The important point for me is that the computer is doing the interpretation not us. We have programmed it to interpret raw data and this is exactly what it does and very efficiently I might say. When the computer takes input, the program processes it and then reacts accordingly. To process the data it must have interpreted the data, that is turn the data into information on the basis of which it acts.

So I cannot see how you can say that a computer doesn't interpret information when this is exactly why they are useful.

The hammer analogy isn't valid, as a hammer has no ability to process or interpret data.

For me the important point is that a computer only ever interprets data the way it is programmed, by contrast we are distinguished from a computer in that we have the ability to change the way we interpret raw data and I describe the process of evolving interpretation: learning. Humans, monkeys, cats, mice all have it, where as computers don't.

Computers have memory, they have the ability to interpret data the way they have been programmed, yet they do not have the ability to evolve their interpretation of the world around them, for this they remain dependant on us via upgrading software.

Kedaman,

quote:
--------------------------------------------------------------------------------

For a start I consider that software and hardware together form a computer. Hardware alone does not constitute a computer, nor does software in isolation.

--------------------------------------------------------------------------------


Assuming this, computers are by definition intelligent.

By your definition of intelligence then yes, and I pointed this out two posts ago.

My definition of intelligence includes the ability to evolve your interpretation of the data received from your environment which I summarise by the word learn. So by my definition of intelligence computers currently are not intelligent as they can only interpret the way they have been programmed and cannot evolve this interpretation.

I know what you want to say even if it's the wrong terminology you use. You mean that AI would start making conclusions on what it percepts and then use these as axioms in a war against us.

Getting very close here!

It's not "learning" that is responsible here, it's the processing part, the intelligence of AI that might overthrow us.

Well I differ here, because to me it is not the processing part that is the problem, but the ability to evolve the processing part (learn) that should never be brought to realisation.

[Nucleus]

Well by my definition of intelligence, which includes the ability to evolve one's interpretation of data, Deep Blue is not intelligent.

Without further upgrades to its programming it will never be able to improve its chess playing ability by itself. On the other hand we can analyse how Deep Blue plays chess and over time alter our own interpretation in order to beat Big Blue.

I would consider Big Blue intelligent only if it could evolve its interpretation of data by itself.

We as programmers give computers the ability to interpret data at a high level relative to many other living creatures such as a dolphin or a cat, yet a humble mouse is eminently more intelligent than a computer. The reason being that at least the mouse was able to evolve to its current level of interpretation and will continue to evolve further, where as the computer requires us to instill it with any ability to interpret data and most importantly has no potential to evolve its interpretation any further.

If you define intelligence only as the ability to turn data into information that is contextually meaningful, and a computer as the combination of hardware and software, then Deep Blue is intelligent.

[simonm]

Nucleus

I do not think computers interpret information. Let me explain.

I don't know why you thought the "hammer" analogy wasn't valid; I didn't say it processed information but anyway, I will give you a closer analogy.

If I have a set of raw data and I draw a graph then I have interpreted the data. I have employed a set of rules that allows me to re-present the data in such a way that it allows my brain to more easilly turn it into information. (Note: it is not information until I read the graph. To someone else who looked at the graph and didn't understand it, it wouldn't be information).

Now, if I transfer my set of rules in to an automated process (such as a software program), the graph can now be automatically generated from the data. Does this mean I have transfered intelligence to the graph generation process? I think not. The program has not interpreted the data, merely re-presented it. It is only by someone looking at the graph, extracting meaning from it and interpreting it that turns it into information. The graph has no meaning to the software program that generated it. It is just an automated process set into motion by it's programmer.

For me the important point is that a computer only ever interprets data the way it is programmed, by contrast we are distinguished from a computer in that we have the ability to change the way we interpret raw data and I describe the process of evolving interpretation: learning. Humans, monkeys, cats, mice all have it, where as computers don't.

I think what you are talking about, an evolving interpretation of data, is just the next level of intelligence. A meta intelligence. Whilst this level of intelligence is always going to be superior to a static interpretation, it is superflous for the basic definition.

I don't think Deep Blue was intelligent. It was just a transfered set of rules. It wasn't even that sophisticated by todays AI standards.

[HarryW]

Nucleus, ¬ is shorthand for boolean 'not'.

Simon, I would be interested to hear what you think humans do with the graph in the process of turning it into information that is radically different from what the computer does.

Data is just numbers or letters, information is data with associated context, and knowledge is information (facts) with associated rules for the inferral of new information. Those are my definitions, and they're what I've been taught (more or less); if you disagree with that, say so.

A graph is placing raw data into a context, is it not? The raw data, though each datum is pretty meaningless on its own, becomes meaningful when given in relation to the rest of the data. That in itself still requires further processing before it becomes knowledge, but I would consider it to be information.

[kedaman]

I don't think Deep Blue was intelligent. It was just a transfered set of rules. It wasn't even that sophisticated by todays AI standards.

I am a bit disappointed at you simon.
Do you think other humans are intelligent?
Why would they not only be your "tools" to interpret data?
What in particular does differ them from Deep Blue, in the aspect of intelligence?

I'm not talking about degree of intelligence, I'm talking about whether they are or not, by definition.

[simonm]

Harry

Data is just numbers or letters, information is data with associated context, and knowledge is information (facts) with associated rules for the inferral of new information. Those are my definitions, and they're what I've been taught (more or less); if you disagree with that, say so.

I would say that I pretty much agree with your definitions acept that data doesn't have to take the form of numbers and letters. It could be anything. This seems like a minor quibble but my reasonning will become clear shortly.

A graph is placing raw data into a context, is it not?

No. I don't think it is. Interpretation places data into context. Raw data can be systematically turned into graphs, it does not require interpretation.

The raw data, though each datum is pretty meaningless on its own, becomes meaningful when given in relation to the rest of the data. That in itself still requires further processing before it becomes knowledge, but I would consider it to be information.

It is meaningful only to someone who knows how to read a graph, who understands what the Axes are and the purpose behind the graph. To someone who doesn't understand the graph, it is not information.

I believe, like beauty, information exists in the eye of the beholder. Something is informaiton to me because it means something to me. It cannot be, objectively speaking, called information because that would be stripping away the context. Information exists only in a relationship between raw data and an entity that can make sense of it.

[simonm]

Kedaman

I am a bit disappointed at you simon.

I seem to be disappointing many people today. Oh well.

Do you think other humans are intelligent?

I believe other humans extract meaning from what they see around them. This, to me, makes them intelligent. I may be mistaken of course but there you have it.

Why would they not only be your "tools" to interpret data?

I have no direct control over them so they are certainly not my tools. I cannot use people to simplify reality in order that I might extract more information from it.

What in particular does differ them from Deep Blue, in the aspect of intelligence?

Deep Blue does not extract meaning from raw data. It processes raw data according to a transfered set of rules. It is an automated process set in motion by it's programmers. If you want to call that intelligence, fine. I don't.

[HarryW]

Simon, you're not explaining how you think the way the computer does this so differently to how a human does it.

I concede the point about numbers or letters, I was just simplifying things a bit.

Raw data can be systematically turned into graphs, it does not require interpretation.

By the time it's in a graph, it has already been interpreted. How is the way a computer creates the graph any different to the way a human creates a graph? I am not saying the computer knows what it means (yet), I am saying that if the context is known to the observer then the data has been manipulated to become information.

Knowledge of the meaning of the data comes at a later stage, and for the record yes I do think a computer can have knowledge of it, with further analysis. This is the way knowledge based systems work. There are numerous examples of them in industry, such as MYCIN that the NHS uses for diagnosis, and there's one DEC uses for quality control of the large and complex computers that it builds for specialist purposes.

It's not human level AI, but it is AI.

[simonm]

Harry

By the time it's in a graph, it has already been interpreted.

No it hasn't!

How can you say that anything exists in it's own right as information? It is only when someone or something (intelligent) reads the graph that it may become information (in the mind of the interpreter). The graph is merely a representation of the data that makes it easier for our brains to make sense of it (i.e. turn it into information).

How is the way a computer creates the graph any different to the way a human creates a graph?

Did I say that a computer creates graphs differently than we do? No I didn't. I said that computer programs that create graphs are just executing a set of rules that we have given them. It is not the generation of the graph that conveys information. It is the interpretation of the graph that is conveys information.

I am not saying the computer knows what it means (yet)

It's not a matter of whether a computer can know what it means, it can't even extract any sort of meaning from it whatsoever. Therefore it is not intelligent.

[HarryW]

It's not a matter of whether a computer can know what it means, it can't even extract any sort of meaning from it whatsoever. Therefore it is not intelligent.

I was not suggesting that intelligence is required to assemble data into information. Nevertheless, I see it as information, since the data has been assembled into a context in which it is meaningful to one who is familiar with the format of a graph. That is information, data in context, I don't agree with you when you say that if somebody doesn't understand it it's not information. I think it is.

You keep saying that computers cannot know anything, that they can't understand things, and generally that there is no meaning to anything they do. I suppose this stems from the fact that they just run through algorithms? Why do you think this is so different from human brain processes? When abstracted to a higher level, you can go beyond algorithms and data, and use heuristics, facts, rules and meta-rules. You can deal with uncertainty.

From what you say it seems like you think computers are devoid of intelligence because they need to have the programs that they run written for them. Is that correct?

[simonm]

Harry

Did you say that you were familiar with information theory?

I was not suggesting that intelligence is required to assemble data into information. Nevertheless, I see it as information, since the data has been assembled into a context in which it is meaningful to one who is familiar with the format of a graph.

You seem to be suggesting that something can exist as information regardless of whether there is an observer there who can extract meaning from it. This is contrary to what information theory suggests. For example, if I said to you: "Your name is Harry". Does that statement contain information? You already know that your name is Harry (and that I already knew that) so it contains no information. To someone who did not know your name was Harry, it does contain information.

The point is, something cannot exist as information in it's own right. You are suggesting that this graph exists as information in an objective sense.

That is information, data in context, I don't agree with you when you say that if somebody doesn't understand it it's not information. I think it is.

To you it is. To the uneducated observer, it isn't (or at least not the same, anyway). It is information only as it exists in your brain. In your brain, it has been interpreted. In the un-educated brain, it has not been interpreted.

I would say, after thinking about it for a while, that a graph is just a coded up version of the raw data. It requires knowlege of the rules that constructed it to enable us to interpret it.

You keep saying that computers cannot know anything, that they can't understand things, and generally that there is no meaning to anything they do. I suppose this stems from the fact that they just run through algorithms?

I do not mean to give the impression that computers will never be able to be intelliegent (or consious), only that a piece of software (such as a graph generation program) is not intelligent.

Why do you think this is so different from human brain processes?

I have thought a little about this and I think that if we are deterministic in nature then we probably aren't intelligent. Perhaps free-will is required to really be intelligent? If I'm just following an algorithmn (however complex and sophisticated it might be) then am I really intelligent? What's so intelligent about just following a pre-set course?

If I have free-will, the ability to choose from a multitude of outcomes (none of which are innevitable), then I can be judged on my intelligence, my ability to make good or bad decisions.

[hypnos]

Consciousness is certainly completely independent from the human body. The human body is simply a vehicle for our consciousness. This explains why people (everybody!) has the ability to shift their consciousness to distant places (i.e. Remote Viewing and Astral Travel etc).

Remote Viewing has been made famous by the American military, which used this to obtain information about enemies in the cold war. They say they stopped because it was too unreliable, but I doubt it.

If you believe in life after death, I’m sure you must believe that the consciousness is an external entity?


This is why; I believe, that a computer can never induce conscious thought.

[simonm]

Consciousness is certainly completely independent from the human body. The human body is simply a vehicle for our consciousness. This explains why people (everybody!) has the ability to shift their consciousness to distant places (i.e. Remote Viewing and Astral Travel etc).

Well, that is a viewpoint held by some but it has certainly not been shown to be the case as conclusively as you suggest that it has.

If we believe that consiousness is seperate from the human body, that certainly spares us from having to explain how such a strange thing could be brought about mechanistically but it also raises many other (and perhaps more difficult) questions.

For example, if the consiousness is not physical in anyway, how (and why) does it interact with the brain attall? Why does it even need a physical body? At what point does it become associated with the body? Where do new ones come from?

I could go on but you get my point. If you would care to shed any light on any of the above questions it would be most illuminating.

[HarryW]

So, given that I don't see much of a possibility of free will existing (as discussed), I guess you wouldn't consider humans intelligent.

Jeez, you set pretty high standards

I didn't say I was familiar with 'information theory', whatever that is.

I think you are talking about a different kind of information to the information I'm thinking of. Information is a step down from knowledge. I'm talking about information in relation to data and knowledge, in terms of how far data has been processed into knowledge, not as a subjective entity. I don't really see the relevance of it to AI if it's only information when you didn't know about it already. That removes the possibility of storing it as part of long-term memory - as soon as you know it it's not information - so it's a pretty useless definition to me. It seems more concerned with the acquisition and transferral of information than the storage of it.

I do not mean to give the impression that computers will never be able to be intelliegent (or consious), only that a piece of software (such as a graph generation program) is not intelligent.

Err... so computers might be intelligent, but software won't be? Of course software won't be intelligent, software doesn't do anything on its own, it has a symbiotic relationship with hardware. When you bring them together you have a computer, so what are you trying to say here?

[hypnos]

Nobody has to prove to you, simonm, that the consciousness is external. I for one don't! If you'd like the true answer, you must find it out for yourself! There are ways of doing this. There are many websites explaining in clear terms how to shift the consciousness from place to place. Here is one good example:

http://home.tower.net.au/~rsb/

I can only show you the door, you have to open it

[Nucleus]

Originally posted by simonm
[B]Nucleus

If I have a set of raw data and I draw a graph then I have interpreted the data. I have employed a set of rules that allows me to re-present the data in such a way that it allows my brain to more easilly turn it into information. (Note: it is not information until I read the graph. To someone else who looked at the graph and didn't understand it, it wouldn't be information).

Now, if I transfer my set of rules in to an automated process (such as a software program), the graph can now be automatically generated from the data. Does this mean I have transfered intelligence to the graph generation process? I think not. The program has not interpreted the data, merely re-presented it. It is only by someone looking at the graph, extracting meaning from it and interpreting it that turns it into information. The graph has no meaning to the software program that generated it. It is just an automated process set into motion by it's programmer.

I don't follow any of the logic in the last two paragraphs. You say "If I have a set of raw data and I draw a graph then I have interpreted the data" further you go on to say that the automatic generation of the graph by a computer "has no meaning to the software program that generated it". From this I deduce that you are saying that the act of drawing the graph is considered interpretation of the data when the entity drawing the graph is able to extract meaning from the graph, and the act of drawing the graph is not considered interpretation of the data when the entity drawing the graph is unable to extract meaning from the graph. Is this really the point you wish to make here

[kedaman]

Simon
You define something is intelligent if it can extract meaning from raw data. You also define something can extract meaning from raw data, if you think so. Consider the usefullness of this definition, for a society where common knowledge is important. Try qualifying something artificially intelligent if everyone thinks differently. Since you think this is all fine, I suspect you don't care of the possibility of AI, you can always deny or accept it's existens if you want, if so, why are you here then?

In case I have missunderstood you, can you explain how you qualify something to be able to extract meaning from raw data?


Hypnos
You can consider consciousness as nonexistent. If something exists, it is physical, since consciousness isn't physical, it doesn't exist.