Black-Box Intelligence

[Ouroboros]

From some discussions with Ssel, we're opening this thread to look into the subject of what intelligence is, and most likely we will discuss AI (Artificial intelligence), NN (Neural networks) maybe even Expert Systems and such. Heck, maybe we can touch the subject of genetic algorithms too!

 

The invitation is open for everyone, and especially our special guest Ssel that seems to have a background and experience in these matters.

 

Hans

[Ssel]

Thank you Hans, I should have guessed you would be the one to start this thread ..

 

I would strongly recommend that we limit this discussion because the goal is really to get to the point where the layman Christian and layman Evolutionist can agree on what constitutes “intelligence”.

 

If we get into all the variety of special detailed concerns of each technical type of intellect, then I’m certain we would never get to the goal. In addition the descriptions and issues that evolved during the technical growth of understanding different intelligence types isn’t really necessary (I hope). Although I hope we get to the point of more exactly defining and explaining that thing people call "consciousness"

 

So, to launch the boat for this journey…

 

ALL intelligence MUST institute the following 3 components;

 

1) Awareness (sensory detections)

..There is not a concrete line one can draw between sensory interpretations and sensory associations. The typical nervous system does a degree of both interpreting and cross associating before the “brain” gets involved. The is actually a requirement for advancement. But is also must be limited so as to not misrepresent was actually detected versus what was interpreted as being detected. The variety and volume of awareness senses add to the over all measure of intelligence.

2) Property associations (understanding) and purpose (inspiration to function)

..This includes spatial and time related associations. The complexity of the associations is one of the measurements of the degree of intelligence. The concern of time associations REQUIRES memory and leads to the realization of cause and effect. The ability to calculate probabilities involved in associations is also included as a complex association tool of any and every “mind”. The requirement to use memory as part of those calculations is an unfortunate secondary consequence. It is unfortunate because it is absolutely needed for higher intelligence but it is not well formed on the base DNA process for development (compared to other developed tools). In layman’s terms the “purpose” is often referred to as the “heart”. This also has very many scriptural ties.

3) Influence/Control

..This constitutes a large part of the goal of any intelligence. Without this goal of having affect on something, not only is it pointless to have intellect, but as it turns out, recursive learning and decision paths can not be formed from experience. This doesn’t exclude neural paths from being formed via DNA construct processes, but advancing beyond what the DNA already knew could not occur. Nothing new could ever be understood as it would have no relevance to measure and awareness could not be tailored. Also the concept of manipulative investigation (even by insects) could never cause learning.

 

 

It is common to think of intelligence as that stuff “inside” and not include the sensors nor the controlled outputs. This leads to a very awkward effort to assemble intelligence. Each of these components help determine the effectiveness of the others. Attempting to measure intelligence independent of sensors, for example, is pointless and presumptuous. The intelligence in the box (the brain) is solely there to handle the specific sensory types attached and relate what specific control outputs (muscles) have any purposeful associations.

 

 

I think it would help to discuss this beginning with both sides of the ID vs. Evolution fence before anything else is added.

 

..any questions / comments ???

[Ouroboros]

Thank you Hans, I should have guessed you would be the one to start this thread ..

Of course. The subject was too interesting not to discuss.

 

I would strongly recommend that we limit this discussion because the goal is really to get to the point where the layman Christian and layman Evolutionist can agree on what constitutes intelligence.

 

If we get into all the variety of special detailed concerns of each technical type of intellect, then I’m certain we would never get to the goal. In addition the descriptions and issues that evolved during the technical growth of understanding different intelligence types isn’t really necessary (I hope). Although I hope we get to the point of more exactly defining and explaining that thing people call "consciousness"

Ok. I'll accept the limits of the scope.

[Amanda]

..any questions / comments ???

 

Hi Ssel... Can I determine if I understand what is going on here... please? I'm sure there are those, like me, that are not as versed in this as you and HanSolo, amongst others. As an interested participant, I'd like to be in the same book, if not even on the same page.

 

It seems that proof for ID by irreducible complexity has been over ridden by the possibility of symbiotic synergistic catalytic effect of two or more chemicals/ingredients, together creating and perpetuating more complexity, and avoiding the 2nd law of thermodynamics. Have I got the general idea? I thought this was basically abiogenesis... anyway.

 

Now we jump to consciousness, which was to be my natural following assertion to ID! I assume that leap was made that this process producing complexity, escaping the 2nd law of thermodynamics, is also what creates consciousness?

 

And I have some questions about awareness/consciousness too, like the different levels of it, yet I don't want to put the cart before the horse. I'd like to understand this prior part of my post first. Thanks.

[Ssel]

... As an interested participant, I'd like to be in the same book, if not even on the same page.

Well, the "book" is the subject of "what constitues intelligence?" and we are on the first page

 

It seems that proof for ID by irreducible complexity has been over ridden by the possibility of symbiotic synergistic catalytic effect of two or more chemicals/ingredients, together creating and perpetuating more complexity, and avoiding the 2nd law of thermodynamics. Have I got the general idea? I thought this was basically abiogenesis... anyway.

The reason this thread was started was to get straight to be relevant question regardless of the detailed efforts to argue over exact microscope issues involved in evolution concerns. The ID vs. Evolutionist argument has been going on for a VERY long time, and has gotten to the point of the issue of IC. The ID proponent is now wanting the evolutionist to demonstrate the resolve to the IC issue. If you look at the posts on pritishd's "Intelligent design" you can see the details of why we now wish to discuss intelligence as a separate issue regardless of any evolutionist or ID issues.

Now we jump to consciousness, which was to be my natural following assertion to ID! I assume that leap was made that this process producing complexity, escaping the 2nd law of thermodynamics, is also what creates consciousness?

Avoid "jumping" ...that's where the faith gets your legs broken.

Consciousness (the REAL thing) is not all that easy to understand and no where close to a ground level jump. This thread will get there with exacting details if those from both sides can follow along for a while.

[Amanda]

Ssel, thanks for your patience, and your clarifications.

[thomas]

 

 

ALL intelligence MUST institute the following 3 components;

 

1) Awareness (sensory detections)

 

2) Property associations (understanding) and purpose (inspiration to function)

 

3) Influence/Control

 

 

Ssel

 

Maybe it will be helpful to look at some specific examples.

 

I have been thinking about simple control systems like those involved in an aircondition or a cruise control in a car. As you know, they operate by a simple registration (what is the current temperatur?, what is the current speed?), and then this is compared with a specified goal (desired temperatur, desired speed), and then action is taken to correct the current situation.

 

Such systems, I believe, cannot qualify as "intelligent", but I think it will be interesting to compare such them with your criteria above. What are such systems lacking, before they can be called "intelligent"?

 

Another questions is how concepts like "reactive" and "proactive" relates to "intelligence".

[Ssel]

In this discussion it is going to be important to try to stay on an abstract level as much as possible. When pointing out those 3 components, my intention was not merely to say that “intelligence” must have these components, but also that “anything whatsoever that has these components has a degree of intelligence.” Let me now explain why.

 

It is often said that some people are intelligent and some are not. What is meant is merely a distinction between degrees of intelligence. Such usage is fine for common language, but in a technical arena such usage becomes a serious problem. When in the technical field vague lines need more solid definitions. World organizations are often formed to help the technical world resolve a globally accepted technical definition so as to allow many countries and businesses to more easily communicate. Sometimes new words have to be created to fill in gaps between common understandings of already accepted words.

 

Is a dog intelligent, or a spider? Well, by human standards, certainly not. But the guiding criteria for the abstract idea of intellect is the ability to solve problems. In this statement the word “problem” refers to anything identified as a situation to be responded to in correction. So, does the spider solve problems? He certainly can’t resolve math problems or resolve any unified field theory, but if you have ever closely examined a spider’s behavior patterns, it becomes amazingly evident that he is a wiz with his own level of problems. So it becomes an issue of degree and also type of intellect and/or problem solving skill.

 

Is an insect intelligent? Does he EVER sense a problem of any kind and respond such as to correct for it? He certainly does. These are often referred to as “defense mechanisms”.

 

Now to the exact question of that A/C control unit. Does it problem solve? I would hope so, that is what it was designed to do. The problem that it detects (awareness) is that the temperature is getting out of acceptable range. But then does it have any complexity at all to its response, any associative connections such as to analyze the situation. Well, it might have a lot more than you realize depending on your knowledge of such things.

 

The control unit for the A/C doesn’t merely go from sensor to output control. Even the simple units use a form of memory to determine where the preferred temperature is to be (the setting). It then uses a probability calculating method to determine that the temperature really is high enough to do something about. It then remembers that it is "in process" of correcting the problem such as to alter it’s responses until such time as the problem has been corrected. All of that was done, in a simple unit, by a hysteresis device which does not flip the switch on until the sensed temperature is slightly above the setting and then once the switch is on, it doesn’t turn it off again until the sensed temperature is slightly below the setting. Of course this is just the simple unit without all of the timers, humidity sensors and such.

 

Obviously the unit has an output control (influence). Thus it has all 3 components and it does respond to solve a type of problem and it happens to solve it in a not so trivial manner.

 

So is it “intelligent”? I must say “yes, it is”. But compared to a human or even a spider? “No, not most humans or spiders.”

 

_______-

 

 

Reactive and Proactive:

 

These relate to a little higher abstract concern and really mostly used in social manipulations.

 

Reactive refers to a control device (or intelligence) that waits until a problem is directly discernable before it response in correction. This is similar to an A/C controller without any timer. Or a person who waits until he gets too many bills in the mail before he finds a better job. Or a manager who waits until there is a problem before he prepares his staff for it.

 

Proactive refers to a control device which anticipates a problem either specifically or in general such that it response in correction before the problem is even detected. This would be an A/C controller WITH a timer. Or a person who finds a job BEFORE he gets hit with bills. Or a manager who anticipates future concerns before they overwhelm his staff.

[thomas]

 

 

Hmmm, if a simple control system can be said to be intelligent, then how about natural selection?

 

This discussion of intelligence started in a tread about Intelligent Design, and the definition of intelligence. And for technical/scientific purposes, a technical accurate definition is of course needed. And although ID basically does not qualify as science (this has been demonstrated in a tread elsewhere on this site), I guess that proponents of ID would like their definition of intelligence to be as scientific as possible.

 

But what kind of definition are ID people currently using? Will they be able to handle a technical definition of intelligence, without jeopardizing their own project? If simple control systems like an air condition can be called intelligent, then how about natural selection? Basically, I believe, natural selection is a control system, although not located in a black box. And if natural selection can be called intelligent, then what’s the point in discussing ID?

 

Of course there are other ways to look at intelligence. I happen to see common language as, although not accurate, the language that best captures all the various aspects of an idea. Technical language at the other hand, is precise and operational, but it also misses some of the finer aspects of an idea. And I have realized, that my common language idea about intelligence involves more, that the technical definition given by Ssel. But of course, for scientific purposes (or for pseudo scientific purposes like ID) a technical definition is clearly needed. But will ID people ever be able to do that?

[Saviourmachine]

Even a more abstract notion of intelligence is possible, using semiotics, the study of signs.

 

The concept "intelligence" in question is related IMHO to "choice". How can somebody make an "intelligent decision" if he had no choice? This is however related to someones belief that choice is an existing phenomenon or that is a human concept that bears no resemblence to the reality (that it only seems that we choose, but it's in fact dictated by our environment). I think "choice" is real, although I don't know yet the analytical basis with their bifurcation points, metasystems, stability, etcetera.

[Saviourmachine]

ALL intelligence MUST institute the following 3 components;

 

1) Awareness (sensory detections)

2) Property associations (understanding) and purpose (inspiration to function)

3) Influence/Control

That seems just a system from Control System theory. With input, a (black) box and output. Adding a feedback loop (environmental changes by the acting components of the box will lead to environment changes in perception by the box) is neat. But, what makes exactly this presentation worthy to be mapped upon the concept of an "intelligent system". Is only this feedback reason enough to call it intelligent is some primitive form?

 

And in Control Systems the function in the box (2) is most often fixed. More general would be to let the function (2) depend from sensory detection (1). But how can such a thing be achieved? How can it without becoming unstable? Or are there more environments that play a role in this game? Or practically separated environments, generated by the sensitivity of the sensory detection systems?

 

I would say, elaborate! I will enjoy some lectures. If you don't mind to spend some time overhere. Thanks!

[Ssel]

Concept 1 (Basic definition of intelligence)

Figure 1 in the attached picture is of the basic intelligence model as discussed earlier.

 

Concept 2 (Feedback and feedforward aspects and models)

Figure 2 and 3 are of the feedback and feedforward abstract models respectively.

 

FEEDBACK

This feedback mechanism is necessary to limit the output influence with regard to a perception input. But this feedback is also responsible for ALL forms of insanity. Inappropriate feedback can be due to a physiological imbalance or disruption or due to a psychological imbalance. We will discuss exactly how this works later.

 

The Buddhist recommends meditation so as to reduce any self-generated output thereby removing the feedback influence. This allows the perception to become more accurate. The concept of meditative prayer serves the same purpose. This effect is present in both the mechanical senses as well as the thought processes which are influenced by their own conclusions. By removing self-generated influence, a perception of reality can become much more accurate.

 

The Magi utilized this natural feedback effect to cause misperceptions of reality and thus produced what came to be called “magic”. The feedback mechanism goes unobserved thus producing an “unseen and mysterious force”.

 

The story of Moses represents a practical application of the concepts of the Buddhist in contest with the manipulation controls of the Magi. We will discuss all of this in greater detail later.

 

 

FEEDFORWARD

Feedforward represents a mechanism of self-observation. The feedforward mechanism can also be used to limit output influence as a feedback of the output before effective influence has been applied to the problem. It can also be used in analysis of the problem and many other forms of learning and problem solving development.

 

The aspect of self-awareness is often inappropriately referred to as “consciousness”. Consciousness is a different higher level process only slightly related to self-awareness. We will discuss the construct of consciousness later.

[Ssel]

But what kind of definition are ID people currently using? Will they be able to handle a technical definition of intelligence, without jeopardizing their own project? ... a technical definition is clearly needed. But will ID people ever be able to do that?

Very good point.

And that is why this topic is being handled in this manner. The point is to propose a usable definition for both sides to agree upon. The ID vs. evolution debate can then continue with more hope of progress.

 

I believe that in this thread, both sides will be able to see the usefullness of this defintion and how it applies. A lot more detailed explanation will be given so as to relate those other aspects that are commonly understood to be a part of intelligence, like emotions, designed purposes for actions and so on.

[Saviourmachine]

I believe that in this thread, both sides will be able to see the usefullness of this defintion and how it applies. A lot more detailed explanation will be given so as to relate those other aspects that are commonly understood to be a part of intelligence, like emotions, designed purposes for actions and so on.

Go ahead. I consider feedback as a necessary but insufficient condition for intelligence. Even for "life" a stronger condition (self-organization) is needed and its questionable if that defines life or is insufficient too. More is needed to define "intelligence" in a bit intelligent way.

[thomas]

Even a more abstract notion of intelligence is possible, using semiotics, the study of signs.

 

The concept "intelligence" in question is related IMHO to "choice".

 

 

 

Recently I became aware of a scientific field, where biological processes are analyzed from a semiotic perspective. As far as I understand, the basic idea is, that in biological systems there is communication going on all the time on many different levels. An organ in the human body for example, is not seen as totally controlled by DNA. The biosemiotic perspective is that the organs reads the DNA, makes an interpretation and then acts on basis of that interpretation.

 

I think a study of intelligence from a perspective involving “interpretation”, will be needed in order to give a satisfying account for the factors involved in intelligence.

 

 

 

Homepage of Jesper Hoffmeyer (a Danish proponet of biosemiotics)

[Ssel]

Even a more abstract notion of intelligence is possible, using semiotics, the study of signs.

The concept "intelligence" in question is related IMHO to "choice".

...

I think a study of intelligence from a perspective involving “interpretation”, will be needed in order to give a satisfying account for the factors involved in intelligence.

What some are you are doing is thinking about the details of what is IN the problem solving box and considering a minimum problem solving skill method or level before it is accepted as truly solving a problem as opposed to merely reacting to stimuli.

 

This first concern in defining intelligence is recognizing that the issue is one of actually being capable of purposely resolving a perceived problem. The point to the basic definition given above is to point out that to resolve a problem, any system MUST be able to detect the situation, have some kind of problem resolving algorithm, and be capable of causing adjustments to the situation. These are the 3 most fundamental building blocks to begin building any complex intelligence system. IF any of the 3 are removed, then the system can not resolve any problems regardless of how complex or great the remaining components might be.

 

The issue of feedback can be INSIDE that problem solving box, but is NOT required as an additional outside component in order to establish a problem solving scheme. The feedforward concept must be outside the box but is only relevant to higher self-aware systems.

 

"Interpretation" will obviously have to come into play in order for the box to identify what it would consider a "problem" to respond to. But again, that is a component INSIDE the box.

 

Our only concern at this level is to agree that the box be able to purposely problem solve once given an input and an output. The exact method of problem solving will determine the type and level of intelligence. In the long run, it is the type and level of intelligence which makes it more readily recognizable as the common understanding of intelligence. Just as in a computer, the most fundamental components in their simplest form is hardly recognizable as a computing device. Yet once those components are made more sophisticated, the computer becomes far more useful and obvious as a computing device.

[Ssel]

Concept 3

Identification and Distribution

 

 

Discrete Identity;

When someone speaks of the intelligence of a computer, they are often referring to the CPU (Central Processing Unit). This component of the computer is generally burdened with the task of associative algorithms which perform math, logical, and routing functions. The CPU box contains many separate IC’s and circuit boards so as to yield a more sophisticated processing complexity.

 

The temptation is to claim that the “intelligence” is merely within the CPU box. And for general usage such would be acceptable. But on a technical and abstract level, one has to ask if merely the CPU box alone contains the required components to yield problem solving potential. That question can be answered easily simply by imagining the effectiveness of a CPU box with no attached wires at all other than the power cord. With no input devices such as the keyboard and mouse and no output devices such as the monitor or printer, can the computer actually do any problem solving? Obviously not.

 

Thus, we must accept that the true intelligence of a computer must include its input and output devices as well as the CPU box. As these sensors and influences change, the type of problem solving (intelligence) must also change. The type and degree of intelligence can not be separated from its sensor and influence potentials.

 

Now think of a human. Where would you be tempted to point if you were trying to point to the intelligence? To the head? If you took the brain out and kept it alive but without its attached nerves, what would you have? It could potentially do property associations but to what end? Can it solve problems? Obviously the analogy is the same as with the CPU. In order to truly identify where the intelligence rests within a human, one must point to almost the entire person. The brain must be attached to sensors and the skeletal neurons must be attached to muscles. The entirety of the muscle mass isn’t required nor any other body components not directly involved in awareness or controlled response.

 

 

Distributed Identity;

Both computers and people are often grouped such as to function together to resolve problems. The awareness component might be from human observation or from directly associated devices. The associative processing of the group is distributed throughout a multi-processing design appropriate to handle the type of problems being addressed. Together the group functions as a team. But these teams must also include an output or influence function without which no problem solving goals could be accomplished. The measure of the awareness, association complexity, and influence potential determine the true measure of the over all intellect and problem solving potential.

 

But regardless of the measure of intellect, what would you point to in order to identify the intelligence? You must point to the entire team. But now what if these computers or people constitute millions of individuals spread over the entire planet? Pointing to the “intelligence” could become a problem.

 

A committee driven group often doesn’t have a “brain” component. Instead the problem solving is accomplished without a central coordinating element. The committee might merely ensure that the group continue making progress without attempting to dictate exact direction.

 

In a church, where is the intelligence? Each element of the church contains some degree of problem solving and the team in total contains a different problem solving potential. The same of true of businesses and other organizations. Mankind as a whole can be said to be a single intelligent entity as long as most of it is in communication with the others. Mankind, as a whole can accomplish what no individual man could even attempt. An ordered structure for all of Mankind allows for problems to be responded to and resolved that possibly no man within was even aware. Mankind becomes a living and thinking entity spread across the Earth handling such problems as environmental changes and immune responses even if no individual was trying to address the issues. Consider which cells within your body have even the slightest idea of your thoughts? Not a single cell with you could even begin to fathom what you are looking at, thinking about, or influencing to any purposeful end.

 

Distributed intelligence allows for thoughts to be developed above the awareness of any member within. The system thoughts can only be identified by watching responses on a higher level then perhaps reverse engineering what it was that triggered all of the responses. The intelligence is above and slightly independent of all members within.

 

It5 might be tempting for a social engineer to claim that there is no intelligence except that of the individuals because he has looked closely and seem nothing but responses of people and machines. Such microscopic vision is common. What the engineer need to identify is the problems that were being addressed even though no one within was attempting such. In doing so, he has identified the entity of Mankind as a thinking being.

[Saviourmachine]

Feedback & Control Systems

The issue of feedback can be INSIDE that problem solving box, but is NOT required as an additional outside component in order to establish a problem solving scheme. The feedforward concept must be outside the box but is only relevant to higher self-aware systems.

What do you mean by feedback in and outside the "problem solving box"? I compare you approach with an ordinary control system. It's possible to formulate a state space description for such systems. It's weird to talk about different feedback or feedforward loops, because such systems use a set of inputs, outputs and states (not just one value). It's also weird to talk about the place of the feedback loop because they can be mathematically / graphically be replaced from one position to the other. For example the line going to box D can also depart from after box B, but the functions in box D have then be recalculated in a standard way. But, you're the expert here, so I suppose you have something different in mind.

 

True intelligence

Thus, we must accept that the true intelligence of a computer must include its input and output devices as well as the CPU box. As these sensors and influences change, the type of problem solving (intelligence) must also change. The type and degree of intelligence can not be separated from its sensor and influence potentials.

I don't agree with that. Everybody knows that if someone disconnects the mind or CPU from its sensors and actuators it can't exhibit its (intellectual) powers. Is this because intelligence is distributed, or does it have another reason? Is the medium (television, radio) important as information carrier, or does it not depend of the way it's received? I would think that it's the latter!

Maybe another approach will lead to what is "intelligent", and what not. Not the approach of detecting interconnections. Interconnectivity occurs in many more systems that "intelligent" systems.

The sensors can be changed like you suppose, but does this really lead to another order of intelligence? Do we want to define "intelligence" as the domain on which we can solve problems or do we want something that is not domain-dependent. Don't we want to say something useful in regard to the difference between the intelligence of a bat (with infrared actuators and sensors) and a human (with visual actuators and sensors)? Intelligence seems a qualitative property, having a scale (IQ) or at least being able to make comparisons. You're definitions of feedback and feedforward systems is not sufficient, there has to be looked inside that box.

 

From "True" Intelligence to Establishing a Map of Intelligence Land

About distributed intelligence and for example that church... I don't think such an analogy applies. When establishing a definition of intelligence we all can agree about, I think it's inappropriate to add questionable "forms of intelligence" such as swarm intelligence, stigmergy, etc. IMHO it's better to use a term like "primitive intelligence" for a definition of intelligence like you purpose. After defining a lot of "intelligence types" we can come up with a common denominator. Or we can define some border that separates "intelligence" from "non-intelligence" (or leave that to someones own imagination). This border is (probably) very subjective. I maybe draw this border between an expert system and a neural network for example.

[Ssel]

After defining a lot of "intelligence types" we can come up with a common denominator. Or we can define some border that separates "intelligence" from "non-intelligence"

Perhaps I'm jumping too far ahead because this is exactly what has already been proposed.

 

The proposition is that a "problem solving" skill be what defines the quality of intelligence and without a problem solving skill the entity has no intelligence.

 

But in addition, which problem solving skills will be effective will depend entirely, and very significantly on which input and output types are available. The bat's brain is designed to problem solve in a dark environment with a different set of inputs than the human's. Thus it's problem solving skills are different, it's algorithms are different, and it's type of intelligence is different.

 

The discussion of exactly which kinds of problem solving algorithms are needed to resolve which kinds of problems is a deeper discussion beyond the basic need within the Evolutionist vs. ID proponent's need to settle on a fundamental understanding of intelligence.

[Saviourmachine]

The proposition is that a "problem solving" skill be what defines the quality of intelligence and without a problem solving skill the entity has no intelligence.

With such a definition I agree. But, I'm also an AI geek.

[Ssel]

With such a definition I agree. But, I'm also an AI geek.

Great

 

So the next and probably MOST significant issue and component of intelligence is that of;

 

Concept 4

Intention

 

The ID proponent not only argues that there is intelligence involved in human creation but more importantly that the intelligence had intent. It is inherent that a problem solving process have intention. It is the intent which defines the goal and the perceived "problem" to be resolved.

 

So, what constitutes intent?

 

Does the A/C controller mentioned earlier have intent? Does a tree's effort to grow have intent? Does a spider's web mastery have intent? Does the DNA replication effort have intent?

 

This thread wasn't really started just for me to lecture on intelligence. But for the lay-Christian and the non-Christian to settle on a foundation understanding such that further debate could take place. I offered to help get the ship out of port so that you each could resolve any differences.

 

Additional concerns of intelligence beyond this concept involve;

 

1) Consciousness - what is it and how does it come about

2) Memory – what constitutes memory and learning

3) Emotion - how does it play into intelligence

4) Can anything solve a problem without having observable influence? – where is God in the picture?

[thomas]

 

Thus, we must accept that the true intelligence of a computer must include its input and output devices as well as the CPU box. As these sensors and influences change, the type of problem solving (intelligence) must also change. The type and degree of intelligence can not be separated from its sensor and influence potentials.

 

 

 

What you are saying is, that intelligence also is a matter of organization. The ability to serve a useful purpose is not only something build into a unit, but also about how the unit is organized together with other units.

 

I have not been thinking about intelligence that way before, but it is very well in line with the insight from system theory, that a system is more that the sum of it's components. The relations between components are important too.

 

In my first response above, I asked about an aircondition. But actually I have an even more simple control system in mind, than an conventional aircondition. I was thinking about a room with a heating system and a piece of metal, that will increase its volume as the temperature goes up. And then at a certain temperature, the metal will reach a volume where is presses a button, that will turn off the heating system. And then the temperature will go down, the metal will decrease in volume, the button will be releazed and the heating system will turn on again. And in this way, the temperature will be keept close to a certain level depending on the position of the metal in relation to the button.

 

Now is this sytems intellligent in its own right? According to the definition "yes", but in my mind "no". Afterall it is only a piece of metal behaving according to the laws of nature. The "only" intelligence I can see, is the intelligence of the person, who has put the system in place.

 

But if looking at intelligence as a result of proper organization, then I don't think I will have problem calling such a simple control system "intelligent" in the sense, that a system has been organized, so a law of nature acts in an intelligent way.

 

But since the purpose involved in intelligence belongs to the user of the control system, I don't think it is intelligent it its own right. But IMHO, it can serve as an extention of human intelligence.

 

 

 

A committee driven group often doesn’t have a “brain” component. Instead the problem solving is accomplished without a central coordinating element. The committee might merely ensure that the group continue making progress without attempting to dictate exact direction.

 

 

 

I once read somthing about "group thinking". I believe it is a proven fact, that a group of people sometimes makes decisions, that no one of the individual members like. And I am not taking about a compromise, where no one of the parts are entirely satisfied, but about groups that makes stupid decisions because some group dynamics influences the process.

 

So a group of intelligent units (people) may together produce unexpected results.

 

 

 

 

 

Distributed intelligence allows for thoughts to be developed above the awareness of any member within. The system thoughts can only be identified by watching responses on a higher level then perhaps reverse engineering what it was that triggered all of the responses. The intelligence is above and slightly independent of all members within.

 

 

 

Doesn't we need to identify a purpose in order to talk about intelligence? And what is the purpose, if we aren't aware of the process? Or will some of the people involved in that process be aware?

 

But nevertheless, from an academic point of view, I find the idea interesting. But what are the political, religious and social implications of it. And why isn't it going on already?

 

I do not know if I am right, but it seems to me that some German philosophical ideas about "Volk geist" (the spirit of the people) are involved here.

[Ssel]

Doesn't we need to identify a purpose in order to talk about intelligence?

Yes, and that is what the "intent" mentioned above is

 

intent = purpose

 

But exactly what constitutes purposeful intent? It is easy to see that a human has purposeful intent. But exactly what initiated that and what defines its exact origin and construct? When is something merely reacting versus reacting with intent or purpose?

[thomas]

Doesn't we need to identify a purpose in order to talk about intelligence?

Yes, and that is what the "intent" mentioned above is

 

intent = purpose

 

But exactly what constitutes purposeful intent? It is easy to see that a human has purposeful intent. But exactly what initiated that and what defines its exact origin and construct? When is something merely reacting versus reacting with intent or purpose?

 

As I see it, purpose is related to humans. We cannot say that the nature has a purpose, the only think we can say about nature is, that it behaves according to certain laws. We cannot identify a purpose behind the laws, and therefore we cannot identify purpose in nature.

 

Therefore technical systems utilizing laws of nature, are not intelligent in their own right, but they can serve as an extension of human intelligence, because the laws of nature can be organized to work according to a human purpose.

 

Also a person "A" may involve a person "B" in a problem solving process, although person "B" does not see or know about the purpose. If person "A" does so, person "B" has become an extension of the intelligence of person "A".

 

But if person "A" and "B" are connected in a process, where nobody knows the purpose, how is it then possible to talk about intelligence? And therefore, how is it possible in a meaningful way to talk about all of humanity as one single intelligent entity?

[Ssel]

As I see it, purpose is related to humans. We cannot say that the nature has a purpose,..

 

.... where nobody knows the purpose, how is it then possible to talk about intelligence?

Two questions, first what does one person knowing the purpose for something(one) else have to do with it having a purpose? You are implying that if a man doesn't know the purpose for something, then it can't have one. Obviously this leads to the idea that only man can have purpose and all else is dictated by him. Is that what your saying? Even God can not have purpose because man didn't get the memo?

 

Secondly, where did the idea that purpose is only related to humans come from? The wolf has no intention or purpose in chasing its prey? The bird has no purpose or intention in building its nest?

 

Granted that the human quite often has a longer chain of related purposes that an animal could not fathom. A man walks into a store and purchases a small bag of wood screws. Rover looks up and watches. He gets back in the car. Rover can't see how the screws had anything to do with helping him get into the car. He drives home and gets a ladder out from the garage. Rover is certain that his master is simply senile for buying the screws that obviously have been forgotten and had no purpose in buying. He climbs up upon the roof with some tools. Rover is annoyed that he can't see what his master is doing. He uses the screws to refasten and repair a pipe at the edge of his house which leads from the water tower down into the house and garden. He comes down off the roof and goes to test the garden valve. Rover is pleased that his master finally remembered to fill his water dish instead of foolishly playing with screws.

 

Very often one purpose is merely to feed another. Thus it is at first difficult to see where the chain began. But it did. To say that this purpose is only to feed that purpose is not answering the question. The question is

"what constitutes intention or purpose?"

 

What is the seed made from? How does any purpose ever get started? This might seem a question for philosophers, but we are beyond that. It is now a question for precise technical analysis. So think very exactly and technically. What causes anything to have intent?