|
Author
|
Topic: Intelligence -- what is it?
|
warren_bergerson
Member
Member # 262
|
posted 26. May 2002 13:31
Mturner,
While the active-passive, dynamic-static distinction is important in understanding intelligence, the ‘knowledge-intelligence’ distinction is more complex than the dynamic-static distinction. To understand intelligence, (IMO) you have to understand how intelligence ‘a property of a set of processes’ can operate on itself to increase intelligence.
To understand this ‘operating on itself’ feature, you need to define intelligence as ‘a property of a set of processes operating to produce purposeful changes in processes’. To understand this concept, the self-programming computer analogy is useful. A computer or logic machine involves 1) occurrences of information processing, 2)programs that define how information is processed, and 3)sets of processes for creating and modifying programs. A program in this analogy is both a static fact and an active process which defines or creates facts or occurrences.
"Real" intelligence, is a property of a true self programming system where the programming process operates on programs, including the programs or processes which control programming. Over time, such a system has the ability to increase both the magnitude and nature or physical manifestations of its intelligence.
The claim that biological systems exhibit intelligence and evolve, is (IMO) the claim that biological systems are ‘intelligent self programming systems. If you can define how such a system actually works, then you have explained both intelligent design and evolutionary change.
A lot of confusion on this issue arises from the claim that Darwinian ‘mutate-select’ systems are intelligent, self programming and capable of evolving different forms of intelligence. My analysis clearly shows that mutate-select systems don’t have anything close to the ability to simulate intelligence. This does not, however, mean that there are not logical sets of processes or systems capable of simulating intelligence.
Bertha,
Choice is clearly a feature of intelligence. But choice is also an indirect feature of fact. A computer program is a fact, but a fact that defines how a choice is made.
IP: Logged
|
|
Berthajane Vandegrift
Member
Member # 272
|
posted 26. May 2002 16:07
Warren wrote:Choice is clearly a feature of intelligence. But choice is also an indirect feature of fact. A computer program is a fact, but a fact that defines how a choice is made.
Hi Warren, A computer program defines how a particular computer makes choices, but I'm not yet convinced of a program that will define how bilogical choices are made. (At least if such a
'program' exists, it hasn't yet been defined.)
If or when we have a deterministic explanation of biological choice, perhaps biological creativity (or choice) could then be considered a
feature of fact, rather than a creative event.
IP: Logged
|
|
mturner
Member
Member # 190
|
posted 26. May 2002 17:26
**
Hi James;
On the distinction between Knowledge and Intelligence; clearly you and I are of very different opinions.
You seem not to distinguish between 'intelligence' and 'knowledge', whereas I have shown why I do see a very important distinction between them. Bertvan has expressed it clearly and succinctly with reference to the act of choice. Intelligence as the act of choosing; knowledge as the product or result of that choice. That is, I believe, the key distinction I was trying to make in my post.
But let me go further and say that Intelligence is the active, intentional- (i.e.,not the accidental or random)- agent for *change*
in Form, whereas Knowledge is the passive 'Form' itself;- that is, some kind of entity; defined, limited, constrained, differentiated, and, to some temporal extent, conserved. Thus Intelligence is about alteration of forms, while Knowledge is about conservation of forms.
I believe that knowledge can be conserved without being applied, simply as a 'form' or 'state of being'. Intelligence is required to create, alter, or amend knowledge, but no intelligence is *required*- (although it may be involved)- in the application of knowledge, and certainly no intelligence is needed to merely store, conserve, or retrieve it. These are mechanical, pre-determined, inflexible, immutable, automatic, reactive, reflexive operations, built into the Form itself. So yes, *stored* Knowledge not only can, but always does, exist apart and disconnected from Intelligence. As soon as it comes in contact with Intelligence, it is no longer *stored*, but is either applied or altered. "Learning" is a major application of Knowledge. [Please note that, as I mentioned above, knowledge can be applied both mechanically, reflexively, and/or intelligently, intentionally (that is, teleologically).]
It is your third paragraph that is antithetical to my POV. BTW, I, too, harbour a deep and abiding fondness for the verb, "to know", as in, "to possess knowledge". But do not confuse "to know" with "to learn". "To learn" is an activity; "to know" is a 'state of being', (that is, 'Form').
I will say again, that I believe you are badly mistaken to conflate Knowledge with Intelligence, and I have tried to show why the two concepts should be kept distinct. I cannot for one moment agree that knowledge ceases to exist as soon as it is no longer being directly accessed by some form of intelligence, as you seem to be saying. I do not believe that the universe ceases to exist the instant that it escapes my attention. I believe that if a tree falls in the forest, it makes a noise, whether anyone hears it or not. In other words, I believe, for all practical intents and purposes at least, in an independently existing 'reality', and as a part of that reality, the independent existence of 'stored knowledge'. I know that there are many who quarrel with this metaphysic, and that there is much room for debate, but I hope we can avoid going into that here.
I hope you have found some validity in my argument to the distinctiveness of intelligence and knowledge, but if not, I look forward to your feedback.
mturner
P.S.
I do not have much free time at the moment, so if I'm slow to respond, please try to understand.
m.
IP: Logged
|
|
mturner
Member
Member # 190
|
posted 26. May 2002 18:43
**
Hi warren_bergerson;
Your first two features of a self-programming computer are irrelevent to intelligence, as they are merely mechanical operations. But "3)sets of processes for creating and modifying programs." is all about intelligence, the nub of the matter. Unfortunately, "sets of processes", doesn't mean much at this stage. It seems to me that you are simply saying that a truly self-programming computer would possess, "3/", the faculty of intelligence. Therefore I can't accept your analogy.
Likewise, your statement about "real" intelligence may true, but, as described here, affords me no insight into the meaning of intelligence. Programs are Knowledge, not Intelligence. When you say, "such a system has the ability to increase both the magnitude and nature or physical manifestations of its intelligence.", you are assuming that it possesses the faculty of intelligence to begin with! Either that or you are saying "intelligence", when you mean "knowledge". It is the nature of the faculty of intelligence, and not what machines could theoretically do with it if they had it, that concerns us here.
Please note that at this stage I have no ideological bias against computers attaining what you term, 'real intelligence', as opposed to their current limitation to an ability to mechanically process and apply knowledge. I simply do not consider what they do now to be 'intelligent', nor am I willing to assume that they can and will aquire 'real intelligence' as a matter of course.
mturner
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 27. May 2002 10:53
Bertha,
Bertha: A computer program defines how a particular computer makes choices, but I'm not yet convinced of a program that will define how bilogical choices are made. (At least if such a 'program' exists, it hasn't yet been defined.) If or when we have a deterministic explanation of biological choice, perhaps biological creativity (or choice) could then be considered a feature of fact, rather than a creative event.
The issue here is a mathematical technique or concept. You are thinking of biological choice as a very complex form of information processing. In effect, you are viewing a decision or choice as being determined by all the input that occurred prior to the choice. I will concede that this type of deterministic algorithm either doesn’t exist or is too complex be identified.
What I am proposing is a different way of looking at the problem. I am suggesting that the complex biological decision making algorithm can be divided into 1)a relatively simple, mechanistic decision making process modeled by a relatively simple algorithm or program, and 2)a more complex process or set of process for selecting the particular ‘simple algorithm or program’ which will be used to make the decision.
In simple notation, this approach suggests that biological decisions are algorithms of the general type f(S)=R where S in stimuli or input and R is output or response. The programming part of the process involves ‘selecting the function f from a set of possible functions F, such that f(S)=R increases the likelihood of achieving some goal G". The fact that f is selected to achieve goal G explains why so many of the features of biological systems are or appear to be teleological or purposeful.
It will be noted that the program/programming sub-division I suggest is 1)a perfectly legitimate analytical technique and 2)there is no logical necessity that the approach needs to work any better than alternative approaches. The technique appears to actually be useful and practical, because it turns out 1)there are actual physical counterparts to the function f, 2)it is both possible and practical to observe, measure and model f, and 3)it is both possible and practical to observe, measure, and model the processes involved in modifying, creating and selecting f.
As might reasonably be expected, one of the most difficult parts of explaining a new approach to the analysis of intelligence is convincing people to try looking at the problem from a new perspective. The logic generally used in analyzing intelligence seems to be 1)the problem must be viewed from this perspective, 2)the problem can not be solved from this perspective, 3)therefore the problem can not be solved.
Mturner,
MT: Intelligence as the act of choosing; knowledge as the product or result of that choice
I am a bit confused as to why you insist on your particular definition of intelligence. Does it lead to some particular type of analysis or conclusion or does it simply your opinion on the subject?
I suggested as an alternative a 3 part concept-1)intelligence as a property of the processes which select or programs to make choices, 2) algorithms or programs which process information and generate choices, and 3)facts or occurrences which are the result of choices. I suggest this alternative concept because my analysis suggests that it leads to a useful understanding of intelligence and how biological systems make choices. As I have stated elsewhere, I stand ready to provide the admittedly complex details of the approach I have developed.
Again, I am probably missing something in what you are trying to say. Could you clarify your comments?
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 27. May 2002 11:44
mturner:
Clearly, a distinction between "knowledge" and "intelligence" is often useful for colloquial purposes. But my view of the fundamental process of cognition does not lead to such a hard and fast distinction as you suggest. So it is my view of cognition that is the source of our disagreement. Let us discuss that for a moment.
All biological functions are fundamentally oscillatory in character. Take an enzyme. It is a structure whose functionality consists in oscillating between conformational substates---not, however, in accordance simply with the laws of physics, butalso in accordance with the overall functional logic of the cell.
Now, my question to you is: Do you consider this power of the enzyme to oscillate in a teleologically appropriate way as "knowlege" or "intelligence"? (It is this appropriateness that constitutes the selectivity or choice that Berthajane refers to, BTW.) It is a pre-existing condition, not "learning." And yet it is an action, and so verb-like. I submit that this (or something like it) is the physical foundation of all cognitive processes in life, and that cognition is above all a form of action. Whether we call it "knowlege" or "intelligence" is no big deal, in my view---it is both, depending on what aspect of it we wish to focus on (instantaneous state, vs. potential capacity). In other words, "intelligence" is not just the capacity to learn, it is the capacity to act in a teleologically appropriate way so as to attain a goal. This capacity surely constitutes "knowledge," as well, in some sense.
Now, you made one other comment that I must respond to, because it represents a fundamental misunderstanding of my position. You said that the world still exists even in the absence of cognitive agents. I agree wholeheartedly! But you imply that "knowledge" must enjoy the same ontological status as the other concrete objects and processes in the world, in that it exists in and of itself, in the absence of cognitive agents. This is begging the question.
On my dynamical view of cognition, knowledge only exists where living processes exist. No living processes, no knowledge. You mustn't (on my view) confuse intrinsic functionality (normativity and semanticity) with externally determined functionality. A book contains "knowlege" (or better, "information") only in relation to a cognitive agent. In the absense of a knower, there is no knowledge, only a bunch of black marks on white paper.
James
[ 27 May 2002, 11:55: Message edited by: James A. Barham ]
IP: Logged
|
|
Berthajane Vandegrift
Member
Member # 272
|
posted 27. May 2002 18:33
Warren Bergerson wrote:
As might reasonably be expected, one of the most difficult parts of explaining a new approach to the analysis of intelligence is convincing people to try looking at the problem from a new perspective. The logic generally used in analyzing intelligence seems to be 1)the problem must be viewed from this perspective, 2)the problem can not be solved from this perspective, 3)therefore the problem can not be solved.
Hi Warren, I agree that unwillingness to look at problems from a new perspective is one of the faults of present day evolutionary theory. I suspect your and my disagreement might be a matter of defintions. In any case, I would never dream of saying anything to discourage you from the kind of analysis you are attempting. I probably have an automatic aversive reaction to anything resembling the Darwinist assertion that "as-yet-undiscovered naturalistic mechanisms" explain evolution.
IP: Logged
|
|
complex
Member
Member # 259
|
posted 28. May 2002 01:17
All,
I am wondering at the ambiguous use of the term "knowledge". Knowledge, in its strong sense, always relates to a truth that exists in or about the world. We can't say for example, "I know that Germany and Japan won World War II". That is a misuse of the term knowledge. If you know something, it must be a truth about the world as it stands. We can either know things empirically discovered or merely grasped conceptually, such as Maxwell's law of electromagnetics. Science often begins with a conceptual understanding before an empirical verification.
In this way, I don't see how the concept of knowledge can be applied to a computer program or mechanism because in this sense it is not a knowing of a truth but a historical state created in response to expternal stimulii. In this low sense of "knowledge" being used in the post, the object in question bears the marks of interaction with the environment. So a ball with a hole in it would have "knowledge" in this sense, in that it preserves a historical mark of some past interaction with its environment. A program that calculates 100 + 100 = 111 (in binary) would have no knowledge of mathematics and in fact gives a false mathematical result -- but with "knowledge" of the falsehood that it was programmed with.
This latter sense of knowledge is not knowledge at all because an interaction with the environment has no truth value, except for the fact of the interaction with external stimulli. The truth interpretation of the interaction cannot be passed on by such interaction.
To *know* a truth about the world is much more fundamental than that. When Descartes writes that he puts all into doubt and finds that he cannot doubt his own self-awareness he is using knowledge in the sense of a truth perceived in a central mind about an issue. He is not talking about a historical mark of interaction with his environment that bears no truth or falsehood value.
IP: Logged
|
|
mturner
Member
Member # 190
|
posted 28. May 2002 02:57
**
warren:
quote:
I am a bit confused as to why you insist on your particular definition of intelligence. Does it lead to some particular type of analysis or conclusion or does it simply your opinion on the subject?
**
I am a bit confused by your question. I am not insisting upon a particular definition, neither of Intelligence nor of Knowledge, but simply offering my opinion. Isn't that what what this forum is about? That is, debate and discussion, rather than lecture and instruction?
As to where my opinion is leading; I would hope to a better understanding of Intelligence. As a tentative conclusion; that Intelligence is a force, like electro-magnetism, gravity, etc., a dynamic agency that expressed through the physical medium. That externally input Intelligence, expressed via the natural laws of physics and chemistry, inform inanimate matter, so that inanimate matter, in all of its various forms, 'knows' how to act, and thus, to be. That living organisms possess more than this exogenously acquired, formal, that is, inherited, Knowledge, but are capable of generating their own change, novelty, alteration of form, thanks to their own endogenous Intelligence.
The root of my opinions lies in my Dynamist metaphysic, such as it is.
mturner
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 28. May 2002 07:50
complex:
You raise a valuable point that we have neglected so far in our discussion: Namely, the standard definition of "knowledge" in the philosophical literature as "justified true belief."
There is a lot to say on this topic, but perhaps this is not the right forum to go on at length in an analytical philosophy vein. Suffice it to say that, in my opinion, "knowledge" in the above sense is intrinsically linked to the concept of representation (i.e., I subscribe to a correspondence theory of truth), which in turn is linked essentially to human language (i.e., I deny that true symbolic representation occurs below the level of language use).
Now, even restricting "knowledge" to the level of linguistic representation, there is still a glaring problem: Namely, "justified true belief" is subject to counterexamples (the so-called "Gettier" examples, and others). The upshot of all of the discussion in the literature, I believe, is that knowledge in this sense is an idealization that cannot exist in the real world. That is, a correspondence may or may not exist ontologically, but epistemically we can never be 100% sure of it. Knowledge in the sense of certainty is unattainable because we are natural systems, not ideal logical abstractions. (For further discussion, see my "A Poincarean Approach to Evolutionary Epistemology," Journal of Social and Biological Structures, 1990, 13: 193--258.)
Does this mean we should eschew use of the word "knowledge" altogether? I don't see why. Our linguistic or explicit knowledge is firmly grounded in our pre-linguistic or tacit knowledge (see Michael Polanyi, Hans Jonas, and others). In some broad sense of "knowing," all living process consists of intelligent interaction with the environment (that is, instrumentally effective, hence information-guided, goal-directed action). Therefore, life just IS knowing, that is its very essence. But knowing in this sense clearly does not imply infallibility!
So, I conclude that the base meaning of "know" must include fallibility, and that certain knowledge is a mere abstract idealization---a sort of perpetual motion machine, or Fountain of Youth, or Eldorado of the epistemologists.
[ 28 May 2002, 07:52: Message edited by: James A. Barham ]
IP: Logged
|
|
complex
Member
Member # 259
|
posted 28. May 2002 21:19
James,
Thank you for the reply. I am not as well versed in the language that you use, which is why I appreciate your comments. I am learning something!
However, if I understand you correctly, you come to a conclusion that to "know" must include fallibility. I think you mean this in the objective sense of the real world, in that there is will be cases in which we claim to "know" something, actually believe it to be the case, but objectively, actually lack a correspondence between our thoughts and the real world.
If I have understood you right (and that is always a big if as I struggle with your language :-) ) then you and I are saying the same thing. Subjectively the person in possession of the thought belief may call it knowledge, but in reality it will not be -- and -- in an objective sense it cannot be called knoweldge.
If you are not distinguishing between a subjecive labelling of knowledge and an objective correspondence and reality, then I have misunderstood you. If you are saying there is no objective observer -- ever -- to be able to distinguish if the thoughts of someone else are subjectively labelled knowledge but objectively are not, then that is a separate claim. But for the latter, as I believe you said, you believe in an objective reality of the world apart from us and therefore must accept objective truths about the world which we can or cannot know.
I think this is an important sub-discussion to intelligence and not tangent to it.
IP: Logged
|
|
Moderator
Administrator
Member # 1
|
posted 29. May 2002 00:23
Just a note that I will be closing this thread down fairly soon. I make a habit of closing down a thread when I think it has reached full maturity.
If you would like to see discussion on a side topic, feel free to open another thread. However, please make sure that the new thread covers NEW ground and makes a positive hypothesis. I don't merely want to see a rehash of this thread.
Thanks.
Mr. Moderator
IP: Logged
|
|
mturner
Member
Member # 190
|
posted 31. May 2002 02:41
**
James;
Before the moderator closes this thread, (and I disagree that it is ready for closure), I must point out that there truly is a difference between intelligence, knowledge, and learning. Any dictionary can make that plain.
Knowledge is that which is learned. Even the simplest organisms can learn, and thus acquire knowledge. Knowledge is not at all limited to human beings, as per your insistence upon "restricting "knowledge" to the level of linguistic representation". You simply cannot do that. It flies in the face of common, everyday experience. Animals learn. Plants learn. All living things learn. What they learn, if retained, is called knowledge. That which empowers learning, that is the ability to learn, the capacity for learning, is called intelligence.
Thus when you say,"In some broad sense of "knowing," all living process consists of intelligent interaction with the environment (that is, instrumentally effective, hence information-guided, goal-directed action). Therefore, life just IS knowing, that is its very essence. " you really should say "learning" where you say "knowing". Knowledge depends upon retention; memory, conditioning, or whatever. Knowledge consists only of that portion of learning which is retained for future use; whether in the mind, the body, a library, a book, or a file. It is not intelligence, (the ability to learn), it is not learning, (the act or process of changing, adapting, becoming knowledgeable, informed). It is simply storage, (and/or assimilation), and/or recovery of learned material.
To me this is obvious. I did not expect debate on it, but rather to simply to carry forward clarification of the concept of intelligence. There is a very strong connection between intelligence and learning, but I suspect that intelligence is more than simply the ability to learn. I believe that Intelligence is that which creates and alters Form, and that forms are what we mistake for Substance.
mturner
IP: Logged
|
|
complex
Member
Member # 259
|
posted 31. May 2002 11:27
mturner,
I appreciate the distinction that you are making and initially when I read it, it seemed to make sense. Of course "knowledge" would be different than "intelligence" because knowledge is innate (once acquired) and intelligence represents some kind of dynmaic usage. (Even if we allow the label of knowledge to low level of "knowledge" that a tree has to reach for the sunlight).
But now, I am not sure. The reason is that if I distiguish two entities/beings by their intelligence (say a 2-year old and a fully grown, mature man), I am really making a distinction based on their ability to understand and be conscious of distinctions and the ability to make new distictions. This ability is tied to what they have learned over their lives.
What has changed between the 5-year old and the fully grown person is not their innate ability but their level of experience, knowledge and wisdom. And that makes one more intelligent than the other.
Similarly, a man is more intelligent than a tree, not simply by the abilities to capture something from the environment and change accordingly but by a greater understanding of the range of implication and subtleties of distictions which allows him to make mental distictions of those environmental conditions (e.g. "that is not sunlight as the tree senses it, but actually a UV tanning machine and I had better avoid long term exposure").
So, intelligence without knowledge seems to be like saying a man without life. Sure, we can have a body there and call it "a man" but is it really?
Perhaps it would help if you contrasted two entities on the intelligence domain and avoid the term knowledge. It seems difficult to do!
[ 31 May 2002, 11:29: Message edited by: complex ]
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 31. May 2002 16:32
mturner:
I certainly never intended to say that knowledge is restricted to human beings. On the contrary, my whole project since 1990 ("A Poincarean Approach to Evolutionary Epistemology," Journal of Social and Biological Structures, 1990, 13: 193--258) has been devoted to analyzing the process of knowing or cognition as an intrinsic aspect of life as such.
Perhaps my reply to one of the discussants above about "truth" was not clear enough. In that posting, I was discussing the mainstream philosophical concept of knowledge as justified true belief, and trying to show where I disagree with it. Obviously, I won't repeat it all again, but let me just say that I basically reject the philosophical concept in favor of a naturalized concept of knowing. There simply is no such thing in the universe as "knowledge" in the sense of certain belief. All belief is always open to error by its very nature. However, we as language users are capable of conceiving what a "truth" relation would be like from without, just in itself (ontologically), even though we can never know for sure from within that one of our own beliefs satisfies that relation (epistemically). That imagined truth relation or correspondence between concepts and states of affairs in the world is, I believe, intrinsically linked to language. In other words, truth is fundamentally linked to symbolic representation, and only language is truly symbolic (I believe that information use in the organism itself is iconic, not symbolic, and is intrinsically linked to its material instantiation).
All of this is discussed at length in the paper I cited above. But the bottom line, for those who don't care about the philosophical literature, is that I totally agree with you that knowing (or cognition or intelligence or learning or whatever we choose to call it) is found throughout the living world. Indeed, I would go further and say that life = knowing (or, more exactly, life = teleology, and teleology = conation + cognition).
[ 31 May 2002, 16:34: Message edited by: James A. Barham ]
IP: Logged
|
|
|
Printer-friendly view of this topic