Register | Sign In


Understanding through Discussion


EvC Forum active members: 65 (9162 total)
4 online now:
Newest Member: popoi
Post Volume: Total: 915,817 Year: 3,074/9,624 Month: 919/1,588 Week: 102/223 Day: 0/13 Hour: 0/0


Thread  Details

Email This Thread
Newer Topic | Older Topic
  
Author Topic:   Why doesn't AI Falsify ID?
aiguy
Inactive Member


Message 31 of 71 (373340)
01-01-2007 3:59 AM
Reply to: Message 30 by Nerd
12-31-2006 8:34 PM


Hi Nerd,
Nerd writes:
I'm getting tired of this thread, but one last comment.
That's too bad. Perhaps some other ID-friendly sort here would like to try and show why AI doesn't falsify ID (or, more precisely, render ID moot - ID is of course always possible, but since chance and necessity can account for complex specified information, we have no reason to postulate anything else).
Nerd writes:
AIGUY:Now, some people think that human minds do not operate according to chance and necessity, but everybody agrees that computers do.
NERD:Wrong. Not everybody agrees with this, I certainly don't. Computers operate under well known physical laws, and it has little to do with chance. Computers are designed to use specific physical laws & properties (e.g., electricity) in a very specific way -- nothing to do with chance or necessity.
I'm afraid you're quite mistaken. Many AI programs like genetic algorithms (as well as other types of programs, like statistical analyses) incorporate randomness. Usually pseudo-random generators are used, but some systems actually employ physical devices that transduce random physical events (like radio static or radioactive decay) into random input for the computer.
So, many AI systems do use chance. And they all use "necessity" (which is simply referring to these deterministic laws of electricity, etc). Chance and necessity - that's how computers work. Just like evolution (and perhaps like brains too).
Nerd writes:
Most importantly, the fact that computers exist has not been shown to be a result of chance or necessity.
It is not controversial how computers come into existence - we already know how that happens. The question is how does the complex form and function of biology come into existence.
ID argues that only intelligent things can design complex machines, and that purely natural processes (i.e. those that consist of only chance and necessity) are incapable of designing complex machines.
I have argued that since computers can design complex machines, and computers operate according to purely natural chance and necessity, then ID has no basis for claiming that natural processes couldn't give to rise to the complex machinery of life.
Your only rebuttal is that computers are themselves designed by humans. However, this argument is a red herring. Nobody thinks that human beings designed life on Earth! So we eliminate human beings as candidates for the explanation of biology. What's left? Well, we know that pure chance and necessity can generate complex designs, so that's the best explanation. Why posit anything else?

Science is not simply reason - it is much less than that. It is reason constrained by empiricism.

This message is a reply to:
 Message 30 by Nerd, posted 12-31-2006 8:34 PM Nerd has replied

Replies to this message:
 Message 32 by Quetzal, posted 01-01-2007 12:20 PM aiguy has replied
 Message 36 by Nerd, posted 01-01-2007 4:54 PM aiguy has replied

  
Quetzal
Member (Idle past 5872 days)
Posts: 3228
Joined: 01-09-2002


Message 32 of 71 (373401)
01-01-2007 12:20 PM
Reply to: Message 31 by aiguy
01-01-2007 3:59 AM


Intelligence Is as Intelligence Does
Hi aiguy! Welcome to EvCForum.
I have very much enjoyed your succinct and well-reasoned argument. In fact, I have enjoyed it so much that I intend to ruthlessly and unashamedly plagiarize it the next time I am confronted with a Dembskiite argument inre specified complexity and the rest of that rot .
One side comment that seems to me to pertain to your OP that has not been discussed yet on this thread is a potential fallacy of equivocation that appears inherent in any argument concerning computer intelligence. To wit: we have but one single example of known "intelligence" (however we operationally define that term) to get on with - i.e., "human intelligence". It strikes me that we may be imposing artificial constraints on what constitutes "intelligence" in the first place. We are, in effect, saying that a computer can only be considered intelligent if it "acts like a human". Since even human behavior can only be predicted stochastically (in the aggregate, and with a large enough sample), why assume that an "artificial intelligence" would be human-like? Indeed, why should we assume that we'd even be able to recognize non-human "intelligence" if confronted by it? Even Alan Turing couldn't get beyond the problem of defining intelligence without relation to "human" intelligence - his "test" was based on the inability of a human judge to distinguish between a human and a computer in a conversation.
I think this point has a profound implication for ID. They too appear to be defining intelligence only in relation to human-scale "intelligence". Admittedly, they provide their putative Designer with "awesome cosmic powers" (albeit stuck in Behe's "itty bitty living space" [/end Aladin mode]). However, I submit they are merely magnifying existing human capabilities - not postulating something beyond human ken (or beyond human intelligence). There is no inherent reason why a computer - which after all in many areas already exceeds the capabilities of an un-augmented human (especially when we're dealing with distributed processing systems or GA, etc) - couldn't develop something that we could neither recognize nor understand, IMO.
I'd be interested in reading your take.

This message is a reply to:
 Message 31 by aiguy, posted 01-01-2007 3:59 AM aiguy has replied

Replies to this message:
 Message 34 by aiguy, posted 01-01-2007 2:11 PM Quetzal has not replied

  
Hyroglyphx
Inactive Member


Message 33 of 71 (373438)
01-01-2007 2:01 PM
Reply to: Message 25 by aiguy
12-31-2006 3:28 PM


Re: Computer models
quote:
Actually, they don't. It would have to be cognizant of itself and others in relation to itself.
This, then, would be your own particular definition for "intelligence". I haven't seen anything in ID literature that lists this quality as a required component for something that is "intelligent". Can you provide a reference?
There is no reference needed, nor do proponents of ID get to monopolize on what constitutes intelligence. Perhaps we, the participants on this thread, should come to some agreement on what intelligence encapsulates. I offer that an intelligence is:
  • self-aware
  • capable of promoting its will
  • capable of emotive responses
  • is able to perform menial to difficult functions
  • is capable of learning beyond its original capacity.
Also, can you say how we can scientifically ascertain that the Intelligent Designer is cognizant of itself and others?
Of course not. If anyone could do that all controversy would end.
quote:
Computers in no way indicate intelligence.
This is an assertion, rather than an argument.
To reiterate, we'd have to come to some sort of consensus on what intelligence actually is. We can go from there. I think I have offered a classical approach as to what intelligence means. Let me know if you have any objections or if you would like to extend that list to possible variables that I have not yet considered.
Your first argument fails because of the reductio argument introduced in the OP (would you say you are only as intelligent as your Designer designed you to be?).
In a sense. I believe, based on the limited knowledge that I've been bestowed, that the Designer allows for a capacity to learn beyond its original means and to inculcate new ideas and emotions. Philosophically, I say that there is definitely a limit. As to the exact line of demarcation, I am incapable of knowing.
Your second argument fails because it is not true - computers do things that nobody "assigns" to them, just as people do things that they've never been "assigned" to do.
What things do they do beyond what they were programmed to do? How far are you going to extend that they can? Are we talking like Sci-Fi, Terminator II, Matrix beliefs where computers will one day out compete us?
I don't see why you are trying to use AI in order to refute ID, when your only source of contention was obviously designed. Its not as if computers have evolved. They required designers to make them come to fruition. Everything else is thus far sophistry.
If you think you have a testable criterion for intelligence that includes people and the Designer of ID theory, but excludes computers, then tell us what it is.
I'm still waiting for the punchline. I'm unclear on what it is you are trying to assert.
So you are saying that neither computers nor human beings are intelligent?
I'm saying that humans are intelligent, but only as much as the Designer wishes. I'm saying that computers mete out mindless functions. It seems that you equate intelligence with, say, the ability to calculate. I think you would agree that a computers calculations are stored in an internal bank. It isn't actually solving anything. It isn't really thinking. It isn't capable of enacting a will or emotion. Think of it this way. When you play against a computer game, it appears that the game is devising a strategy, when in reality, all its doing is calculating your move with a countermove. The developers made it that way to give the appearance that you are playing an intelligent being. But you're obviously not.
Its the same with Deep Blue which is inarguably an impressive machine, but it is not capable of thinking. If you disagree, explain why that is.
I am offering no particular definition for "intelligence" at all.
I think in all fairness that you should. If you are a proponent of Artificial Intelligence, its critical for you to define what intelligence is in order for us to understand what you are talking about.
I believe the word is scientifically useless, in that while it can be used to loosely describe some sorts of behavior, it cannot be used to explain anything.
Intelligence can't be used to explain anything? It explains the difference between intent and capriciousness.
I object to ID because it claims to be an explanatory theory that offers "intelligence" as its sole explanatory concept, but then fails to provide a standard technical definition for that term.
In one word, Teleology. How can you honestly asset that they haven't made propositions as to what intelligence is? You can say that you disagree with the premise, but I don't think you can say that nothing has been proposed.
I think it's a bit odd for us here on this forum to begin making up our own definition for this word, given there is supposed to be a scientific theory that explains biological complexity by invoking "intelligence".
We make up definitions all the time AIguy. Its called, "language." Every one of those words were assigned meaning. You can't remain ambiguous like this indefinitely and expect for anyone to extrapolate some profound meaning.
Aside from that, your attempt at a definition is not too good: First, there is no scientific test for "will" (but you may be interested to learn about scientific investigations into free will
Maybe that's your problem. Maybe you see everything in a stale, mechanistic manner when you might find more benefit by looking beyond it. I'm not advocating that you jettison science and take up some wholly metaphysical stance on everything. What I am saying is that because you can't quantify will and emotion, doesn't mean they don't exist, or if they do, somehow they are inconsequential.
Second, you haven't supported your assertion that intelligence requires emotion.
You haven't supported that it does not. Let me ask you something. Do we not categorize creatures, as far as it relates to intelligence, by their cognitive abilities? What is the similarity between them all? Answering honestly will invariably lead you to answer your own question.
Third, why do you think learning is essential for intelligence. Do you believe that the Designer learns - or doesn't He perhaps know everything already?
You, perhaps, confuse knowledge with wisdom. And you keep asking everyone else to answer what intelligence means. You obviously have some incredulity that intelligence incorporates will and emotion, or vice versa, but refuse to answer it yourself. That may be a safe haven for you, but its a slippery slope argument.
Most AI researchers (including me) believe that human-like cognition requires much more complex physical interactions with the world that has been attempted to date. In other words, we won't really build a realistic aritificial mind until we also build a more realistic artificial body.
What does interaction have to do with anything if will and emotion are meaningless terms? You can't have one without the other and still possess anything meaningful.

"A man can no more diminish God's glory by refusing to worship Him than a lunatic can put out the sun by scribbling the word, 'darkness' on the walls of his cell." -C.S. Lewis

This message is a reply to:
 Message 25 by aiguy, posted 12-31-2006 3:28 PM aiguy has replied

Replies to this message:
 Message 35 by aiguy, posted 01-01-2007 3:06 PM Hyroglyphx has replied

  
aiguy
Inactive Member


Message 34 of 71 (373441)
01-01-2007 2:11 PM
Reply to: Message 32 by Quetzal
01-01-2007 12:20 PM


Re: Intelligence Is as Intelligence Does
Hi Quetzal,
Quetzal writes:
Hi aiguy! Welcome to EvCForum. I have very much enjoyed your succinct and well-reasoned argument. In fact, I have enjoyed it so much that I intend to ruthlessly and unashamedly plagiarize it the next time I am confronted with a Dembskiite argument inre specified complexity and the rest of that rot.
Please do. The odd thing is that I have yet to engage somebody on these any of these forums who is willing to defend Dembski-style ID. I've met out-and-out creationists, various flavors of vitalists and animists... and mainly those who just like God and don't like science and refuse to engage in actual debate.
You made very good points; here are my thoughts:
The enterprise of AI proper (the computer science discipline, as opposed to philosophy of mind or cognitive science) is quite explicitly aimed at reproducing human mental abilities. There is no pretense of establishing a general theory of intelligence, and that is not our goal. As you point out, Turing recognized "intelligence" is not a property that can be objectively defined and tested, so AI's operational definition is simply "those sorts of behaviors that people tend to call 'intelligent'". So while AI is not actually restricted to human intelligence (e.g. there is a lot of research into insect-like intelligence), we don't pretend that "intelligence" is anything other than a loose, subjective, informal way to categorize certain types of behaviors that strike people in general as being clever.
The enterprise of ID, in contrast, is built upon a monumental equivocation. Clearly, if one takes Dembski's basic definition of "intelligence" seriously, ID is perfectly vacuous: ID explains specified complexity by "intelligence", which is scientifically defined as "that which can generate specified complexity".
But by using the word "intelligence", which is laden with all sorts of folk-psychological connotations relating to human mentality, ID sneaks in any number of implicit and anthropomorphic assumptions that nobody could possibly support empirically. So while Dembski fills books with completely irrelevant mathematics to make it seem that we can scientifically detect something, that something is assumed by the unwashed masses to be a human-like mind. Free will, consciousness, emotion, empathy, intent... all sorts of these intuitive or philosophical concepts are what people typically associate with "intelligence".
So the equivocation of the century is for ID to attempt to empirically support one meaning of the term (which is in fact tautologically, necessarily true of evolution or any process that can create complex designs) while pretending that all of these other distinctly non-scientific meanings of the word are somehow scientifically warranted as well.

Science is not simply reason - it is much less than that. It is reason constrained by empiricism.

This message is a reply to:
 Message 32 by Quetzal, posted 01-01-2007 12:20 PM Quetzal has not replied

  
aiguy
Inactive Member


Message 35 of 71 (373454)
01-01-2007 3:06 PM
Reply to: Message 33 by Hyroglyphx
01-01-2007 2:01 PM


Re: Computer models
Hi Nemesis,
Nemesis writes:
There is no reference needed, nor do proponents of ID get to monopolize on what constitutes intelligence. Perhaps we, the participants on this thread, should come to some agreement on what intelligence encapsulates. I offer that an intelligence is:
* self-aware
* capable of promoting its will
* capable of emotive responses
* is able to perform menial to difficult functions
* is capable of learning beyond its original capacity.
I have no trouble with people ruminating about minds and what intelligence may or may not be. I am only interested here in discussing what people attempt to pass off as some sort of science. If ID does not position itself as a scientific theory and attempt to be taken seriously as an alternative to evolutionary theory, then I say have a nice time thinking about these things.
Nemesis writes:
To reiterate, we'd have to come to some sort of consensus on what intelligence actually is. We can go from there. I think I have offered a classical approach as to what intelligence means. Let me know if you have any objections or if you would like to extend that list to possible variables that I have not yet considered.
There really is no "classical" approach at all. However, I think all of your characteristics are off the mark. In my personal opinion: Self-awareness should be excluded, since this has more to do with sentience or consciousness than intelligence. I would also exclude "will", since the subject of free will can't be resolved scientifically. I would also exclude "emotion", since while humans seem to use emotion in their thinking, not all intelligent agents necessarily would. I would say "menial to difficult functions" has no meaning of sufficient precision, and "learning" is non-essential too, since an ultimately intelligent thing would have no need to learn anything.
But the point is this: We could argue all day about what does or does not constitute intelligence, but we wouldn't be arguing about anything but the meaning of a word. There is simply no scientific fact of the matter.
Nemesis writes:
In a sense. I believe, based on the limited knowledge that I've been bestowed, that the Designer allows for a capacity to learn beyond its original means and to inculcate new ideas and emotions. Philosophically, I say that there is definitely a limit. As to the exact line of demarcation, I am incapable of knowing.
What things do they do beyond what they were programmed to do? How far are you going to extend that they can? Are we talking like Sci-Fi, Terminator II, Matrix beliefs where computers will one day out compete us?
Currently, AI systems reason, make inferences, learn, re-program themselves, and deal with situations unanticipated by their programmers. It is also true that they are brittle: They generally lack the ability to use common-sense reasoning, and are intelligent only within constrained domains. So, while we have nothing approaching the sci-fi versions of AI, neither is there any evidence of some sort of hard limit as to what mental abilities might be replicated artificially.
Nemesis writes:
I don't see why you are trying to use AI in order to refute ID, when your only source of contention was obviously designed. Its not as if computers have evolved. They required designers to make them come to fruition. Everything else is thus far sophistry.
Let me try to make this point more clearly. Either "intelligence" is a property that can be ascertained to exist in an entity, or it isn't. If it isn't, then ID is a big waste of time. If it is, then entity X is either intelligent or it is not.
If X is intelligent, then it makes no difference how X became intelligent. Indeed, when ID critics ask IDists "Who designed the designer?", ID's response is that the ultimate cause of the Designer's intelligence is irrelevant! Therefore, when we look at a computer that does intelligent things, according to ID, we can decide that it is intelligent per se, without considering its ultimate cause.
Given we've established that computers are bona-fide intelligent agents, it is irrelevant that they were designed by humans. Again: If you deny computers are intelligent simply because they were designed, you must also deny that human beings are intelligent if you also believe that we were designed.
Now, what does this prove? It proves that processes that act strictly according to deterministic natural law, combined with random chance, can be intelligent. Both AI computers, and evolutionary processes, act strictly in accordance with deterministic law plus chance. (Most cogntive scientists believe that human minds also work this way, but that is by no means settled science).
The fact that chance plus necessity can yield intelligent behavior does not prove that we were not intelligently designed. It does, however, render the arguments of ID moot. ID claims that something that does not act according to purely natural chance and necessity is required to build the complex machinery of life, but since we have observable proof that nothing but chance and necessity is required for intelligence, ID's claim is unwarranted.
Nemesis writes:
I'm saying that humans are intelligent, but only as much as the Designer wishes. I'm saying that computers mete out mindless functions. It seems that you equate intelligence with, say, the ability to calculate.
I have not equated "intelligence" with anything. I have said there is no scientifically useful definition for the word. For any definition you may care to offer, there is none that is both empirically testable and capable of distinguishing human behavior and the inferred abilities of the Intelligent Designer from things like computers. Therefore, there is no scientific justification for labelling the cause of biological complexity as "intelligent".
Nemesis writes:
I think you would agree that a computers calculations are stored in an internal bank. It isn't actually solving anything. It isn't really thinking. It isn't capable of enacting a will or emotion. Think of it this way. When you play against a computer game, it appears that the game is devising a strategy, when in reality, all its doing is calculating your move with a countermove. The developers made it that way to give the appearance that you are playing an intelligent being. But you're obviously not.
This is a very naive view of AI. No, calculations are most certainly not "stored in an internal bank" (think about it - there are an infinite number of calculations that can be done in simple arithmetic). Rather, AI computers are "born" with some knowledge and abilities, and given additional knowledge by their programmers and by learning.
Nemesis writes:
Its the same with Deep Blue which is inarguably an impressive machine, but it is not capable of thinking. If you disagree, explain why that is.
As my colleague Drew McDermott memorably quipped, "Saying Deep Blue doesn't really think about chess is like saying an airplane doesn't really fly because it doesn't flap its wings."
Nemesis writes:
I think in all fairness that you should. If you are a proponent of Artificial Intelligence, its critical for you to define what intelligence is in order for us to understand what you are talking about.
AI has a very clear operational defintion of "intelligence": It is whatever people might generally tend to call intelligence. That's it. It doesn't make any difference at all what that might be. Nothing changes if you decide to call chess-playing, or disease-diagnosing, or theorem-proving, or language-using,... intelligent or not.
Nemesis writes:
Intelligence can't be used to explain anything? It explains the difference between intent and capriciousness.
Again, please remember I am talking about scientific explanation, not causual conversation, philosophy, or theology. There is no scientific way to detect capriciousness.
Nemesis writes:
AIGUY:I object to ID because it claims to be an explanatory theory that offers "intelligence" as its sole explanatory concept, but then fails to provide a standard technical definition for that term.
NEMESIS: In one word, Teleology. How can you honestly asset that they haven't made propositions as to what intelligence is? You can say that you disagree with the premise, but I don't think you can say that nothing has been proposed.
If you choose "teleology" as your definition of intelligence, that's fine. As Norbert Wiener, the father of cybernetics, defined "teleology" about fifty years ago, it is the property of a system to work towards a goal. Weiner also demonstrated that any system that incorporated negative feedback was teleological. Now, a thermostat is a device that works toward a goal (it tries to keep a room at a constant temperature) and incorporates negative feedback to do so. Thus, scientifically speaking, a thermostat is teleological.
And so, scientifically speaking, when you say the Designer must have been "intelligent" - meaning teleological - you are saying no more than it has the properties of a thermostat.
Nemesis writes:
We make up definitions all the time AIguy. Its called, "language." Every one of those words were assigned meaning. You can't remain ambiguous like this indefinitely and expect for anyone to extrapolate some profound meaning.
Again, you may talk all you like of course, and use or make up all sorts of definitions. However, in order to do science, your definitions must be grounded in emprically observable data (i.e. they must be operationalized).
Nemesis writes:
Maybe that's your problem. Maybe you see everything in a stale, mechanistic manner when you might find more benefit by looking beyond it. I'm not advocating that you jettison science and take up some wholly metaphysical stance on everything. What I am saying is that because you can't quantify will and emotion, doesn't mean they don't exist, or if they do, somehow they are inconsequential.
Please, Nemesis, try to understand my stance here. I of course do not believe that science can capture all truth (look at my signature line!). I will even tell you that my metaphysics are distinctly mystical! I actually happen to believe that our conscious awareness is something that transcends mechanism. In another context, I would happily discuss the merits of dualism, how quantum mechanics may or may not provide a way to understand free will, and so on.
The only point I am making here is that Intelligent Design fails as a scientific theory. That is all.
Nemesis writes:
Let me ask you something. Do we not categorize creatures, as far as it relates to intelligence, by their cognitive abilities? What is the similarity between them all? Answering honestly will invariably lead you to answer your own question.
My honest answer is that there is no way to objectively describe what the similarity is between all "intelligent" creatures. Deciding whether or not some critter is intelligent or not is simply an exercise in definition, rather than discovery.
Nemesis writes:
You, perhaps, confuse knowledge with wisdom. And you keep asking everyone else to answer what intelligence means. You obviously have some incredulity that intelligence incorporates will and emotion, or vice versa, but refuse to answer it yourself. That may be a safe haven for you, but its a slippery slope argument.
I have told you exactly what I think "intelligence" is: A subjective, informal way we refer to some sorts of behaviors. Like "athletic" is a loose way to describe some sorts of physical abilities, or "beautiful" is a loose way to describe some sorts of physical features. None of these words are scientifically useful.
Nemesis writes:
What does interaction have to do with anything if will and emotion are meaningless terms? You can't have one without the other and still possess anything meaningful.
I have never said they are meaningless, I have only said they are generally scientifically untestable. You can provide operational defintions for mentalistic terms in order to make them scientifically useful, but once you do that, you have lost the meaning that you wanted to think about in the first place (like I did with "teleology" above).
Edited by aiguy, : No reason given.

Science is not simply reason - it is much less than that. It is reason constrained by empiricism.

This message is a reply to:
 Message 33 by Hyroglyphx, posted 01-01-2007 2:01 PM Hyroglyphx has replied

Replies to this message:
 Message 40 by Hyroglyphx, posted 01-02-2007 2:12 PM aiguy has replied

  
Nerd
Inactive Member


Message 36 of 71 (373474)
01-01-2007 4:54 PM
Reply to: Message 31 by aiguy
01-01-2007 3:59 AM


OK, can't help another comment:
aiguy writes:
So, many AI systems do use chance.
Of course, I agree. This does not mean, however, that AI results from chance and necessity alone. That is a huge leap which you have not shown any reason to make, and which I assert is completely false.
It appears you are making a mistake that is quite easy to make. You assume that if a system has a random input, then any output from that system is a result of only "chance and necessity".
Please consider this simple thought experiment that proves the above is a false assumption.
Suppose you have a die (as in dice) which when rolled generates completely random numbers from 1 through 6. You have also created a "game" system that has as an input 6 buttons numbered 1 through 6, and a single output, a visual display. The system outputs a different pre-defined picture on the display for each button pressed. The way the game works is you roll the die, push the button corresponding to the number you rolled, and examine the resulting output.
In the above system, the output cannot be predicted because you don't know which picture is going to come up next, depending on the die roll.
Does this mean the output is only a result of chance and necessity? Of course not. The output is a result of chance and physics and design. The game system was designed. The pictures were pre-defined (drawn by some artist).
This system could not exist apart from the design that went into it. To prove that it operates only on chance and necessity, you would also need to prove that the design process itself was only a result of chance & necessity, and that the artist produced the pictures only through chance & necessity.
This same argument goes for any kind of function or algorithm -- just because they have a random input does not change the fact that the input is manipulated through a process that was designed. Thus, the output does not exist apart from that design process.
As an aside: Truth does not care about the personal beliefs of the scientist or philosophist attempting to determine that truth. No single origins theory of any kind has a monopoly on science -- they all leave as many questions unanswered as they answer. Science is a process, not a theory.

This message is a reply to:
 Message 31 by aiguy, posted 01-01-2007 3:59 AM aiguy has replied

Replies to this message:
 Message 37 by aiguy, posted 01-01-2007 6:32 PM Nerd has not replied

  
aiguy
Inactive Member


Message 37 of 71 (373490)
01-01-2007 6:32 PM
Reply to: Message 36 by Nerd
01-01-2007 4:54 PM


Hi Nerd,
Nerd writes:
Of course, I agree. This does not mean, however, that AI results from chance and necessity alone. That is a huge leap which you have not shown any reason to make, and which I assert is completely false.
It appears you are making a mistake that is quite easy to make. You assume that if a system has a random input, then any output from that system is a result of only "chance and necessity".
No, I haven't assumed this for any system, only computers. It may be true of human "systems", but it may not be.
Nerd writes:
Please consider this simple thought experiment that proves the above is a false assumption.
I love thought experiments.
Nerd writes:
Suppose you have a die (as in dice) which when rolled generates completely random numbers from 1 through 6. You have also created a "game" system that has as an input 6 buttons numbered 1 through 6, and a single output, a visual display. The system outputs a different pre-defined picture on the display for each button pressed. The way the game works is you roll the die, push the button corresponding to the number you rolled, and examine the resulting output.
In the above system, the output cannot be predicted because you don't know which picture is going to come up next, depending on the die roll.
Does this mean the output is only a result of chance and necessity? Of course not. The output is a result of chance and physics and design. The game system was designed. The pictures were pre-defined (drawn by some artist).
I'm trying to understand what you are getting at here. I understand what you mean by chance (although randomness is in itself a pretty tricky concept, but I think we can avoid that in our discussion here). I also understand what you mean by physics: As far as physics is concerned, everything is a caused by a combination of chance and deterministic physical law. But then you say there is third type of cause involved, which is design. In other words, you seem to take it as evident that there is something called "design" which does not operate according to the laws of physics. But previously you had said that everything operates according to physical law ("Computers operate under the physical laws of the universe, just like everything else does"). So I really do not know what you mean.
Nerd writes:
This system could not exist apart from the design that went into it. To prove that it operates only on chance and necessity, you would also need to prove that the design process itself was only a result of chance & necessity, and that the artist produced the pictures only through chance & necessity.
Again, I will not claim that human minds can be reduced to physics (i.e. chance & necessity). It may be true, or it may not. So, you may reasonably believe that something outside of physics is going on inside of people that enables them to think, and we can't resolve that currently by appeal to science.
Here is where we are disagreeing, I think: You insist on considering the origin of something, rather than just the thing-in-itself, when you decide whether something is intelligent or not, or whether something operates according to physics or not. My argument, however, is that this consideration is illicit. Entity X is either intelligent or not, and X either operates according to physics or not, and this is true no matter how X came to exist in the first place. If you wish to say that X is not intelligent because something else designed it, or that X does not operate according to physics because something else that may not be constrained by physical law created it, then you are courting a number of paradoxes. The paradox that I have used to make this point is that in your way of thinking, if you believe human beings were designed, then that means human beings must not be intelligent.
Nerd writes:
This same argument goes for any kind of function or algorithm -- just because they have a random input does not change the fact that the input is manipulated through a process that was designed. Thus, the output does not exist apart from that design process.
Again, I think your argument turns out to be an attempt to characterize ultimate cause, which is outside the realm of science entirely. Let me repeat a point I just made to Nemesis, above:
If X is intelligent, then it makes no difference how X became intelligent. Indeed, when ID critics ask IDists "Who designed the designer?", ID's response is that the ultimate cause of the Designer's intelligence is irrelevant! Therefore, when we look at a computer that does intelligent things, according to ID, we can decide that it is intelligent per se, without considering its ultimate cause.
The same is true of "operates according to chance and necessity". We may not know if some mind was responsible for creating the universe, but we understand the universe in the physical sciences in terms of chance and necessity, and (currently) nothing else. You could argue that something non-physical had to design atoms and molecules and gravity, too, but that does not mean that our study of all physical systems is really the study of intelligent causation - it isn't.
Nerd writes:
As an aside: Truth does not care about the personal beliefs of the scientist or philosophist attempting to determine that truth. No single origins theory of any kind has a monopoly on science -- they all leave as many questions unanswered as they answer. Science is a process, not a theory.
Absolutely true - I couldn't agree with you more.

Science is not simply reason - it is much less than that. It is reason constrained by empiricism.

This message is a reply to:
 Message 36 by Nerd, posted 01-01-2007 4:54 PM Nerd has not replied

  
TheMystic
Inactive Member


Message 38 of 71 (373642)
01-02-2007 8:45 AM
Reply to: Message 1 by aiguy
12-30-2006 5:08 AM


I'm not sure exactly how to put this succinctly, but to observe that a higher quality can be fully expressed by a lower medium, yet not comprehended by that medium. Take an audio CD, for instance. If you knew nothing of music or CDs you might analyze such a disc and get as far as finding frequency and amplitude in the pits. But you would never find beauty or sadness, or whatever other emotion the composer of the music was trying to express. You could also note that Beethoven's 5th could be expressed as sheet music, as a CD, or as a live performance. The beauty or sadness can be carried by many different media. So one has to ask if intelligence is a physical quality at all. I think not. Relative to the creation/ID debate I think the important phrase here is "appropriately constructed [computer systems]". You'll really be on to something when you find an intelligent computer system that wasn't constructed.

This message is a reply to:
 Message 1 by aiguy, posted 12-30-2006 5:08 AM aiguy has replied

Replies to this message:
 Message 39 by aiguy, posted 01-02-2007 1:57 PM TheMystic has replied

  
aiguy
Inactive Member


Message 39 of 71 (373717)
01-02-2007 1:57 PM
Reply to: Message 38 by TheMystic
01-02-2007 8:45 AM


Hi Mystic,
Mystic writes:
You'll really be on to something when you find an intelligent computer system that wasn't constructed.
I presume that by "constructed", you mean "designed by human beings". Well, yes, there is an intelligent computer system that wasn't constructed... but we have to define our terms here. By "intelligent" I mean "capable of creating complex form and function". And by "computer", I mean "process that operates soley according to chance and necessity.
So, what is this intelligent computer that wasn't designed by humans? Evolution, of course.
Allow me to anticipate your objections:
Objection: Evolution isn't intelligent! It is a blind, mindless process!
Reponse: By the definition I just gave, evolution is intelligent (capable of creating complex form and function). If you don't like this definition, you can say evolution isn't intelligent... but it doesn't change anything - evolution is still capable of creating complex form and function.
Objection: But evolution can not create complex form and function (CFF)!
Response: Even if the currently understood mechanisms of evolutionary theory fail to account for all biological phenomena, we know that pure chance and necessity can indeed create CFF, since computers do it all the time.
Objection: But if neo-Darwinism fails to account for biology, how do you know that nothing but chance and necessity were still responsible?
Response: Because chance and necessity is the basis for our understanding of everything in the universe - we know of no other type of cause. And we know that chance and necessity can create CFF, so there is no reason to posit anything else.
Objection: But computers are designed by people!
Response: Yes, but evolution was not. Both operate according to purely natural deterministic law plus chance.
Objection: But computers can only create CFF because they were designed by people!
Response: Once they are created, computers can create CFF. And they create CFF all by themselves, without guidance, by operating according to pure chance and necessity.
Objection: But then evolution must have been designed by something intelligent!
Response: Maybe so, maybe not. I won't argue that point. You can believe that evolution - and everything else in the universe - was created by something intelligent, and it doesn't change science in the least.

Science is not simply reason - it is much less than that. It is reason constrained by empiricism.

This message is a reply to:
 Message 38 by TheMystic, posted 01-02-2007 8:45 AM TheMystic has replied

Replies to this message:
 Message 41 by TheMystic, posted 01-02-2007 2:13 PM aiguy has replied

  
Hyroglyphx
Inactive Member


Message 40 of 71 (373725)
01-02-2007 2:12 PM
Reply to: Message 35 by aiguy
01-01-2007 3:06 PM


Re: Computer models
I started to respond to this post yesterday and was nearly finished. I got up to use the restroom and came back to a blank page. Apparently my daughter hit the back button. Yeah, that's frustrating.
Hopefully I can get through the whole post this time.
I have no trouble with people ruminating about minds and what intelligence may or may not be. I am only interested here in discussing what people attempt to pass off as some sort of science.
Fair enough. I was unclear if your thread was meant to speak more about AI or more about ID.
There really is no "classical" approach at all.
If there isn't a classical approach, then how is it that we humans generally agree on the same descriptions concerning intelligence?
Self-awareness should be excluded, since this has more to do with sentience or consciousness than intelligence.
After more careful thought, I agree and retract my previous assertion about self-awareness because it does not necessarily mean intelligence. For instance, an infant may not be aware of the self, yet clearly they are capable of using cognitive reasoning-- something I would say is a clear marker for intelligence.
I would also exclude "will", since the subject of free will can't be resolved scientifically.
Just because science cannot, as of yet, understand what the will is does not negate the existence or that it is somehow inconsequential to intelligence. Would you agree that only intelligent beings are capable of asserting their will? I only ask because it seems that computers perform the tasks assigned to them and only those tasks. They don't exhibit any kind of freewill, whereas, natural beings enact their will at their discretion. I suppose one could make the argument that they too could only be performing the functions assigned to them. For face value I could agree with that. But even if this is so, then natural beings are still far more complex than even supercomputers. But even more than that, it says that those creatures were indeed designed with specifications.
I would also exclude "emotion", since while humans seem to use emotion in their thinking, not all intelligent agents necessarily would.
But this is the very thing that presents a problem for the secularist argument. For a secularist, biology is the only acceptable answer to the question, which invariably means that it is the only thing that offer any answers. An evolutionist will note that a sense of morality or the arts must have derived through natural means. If emotion is separate from intellect, then what inexplicable reason made it so? Where does emotion come from, and is it truly exclusive from intelligence?
I would say "menial to difficult functions" has no meaning of sufficient precision, and "learning" is non-essential too, since an ultimately intelligent thing would have no need to learn anything.
Learning is inconsequential to intelligence? Surely this is incorrect. The greater the intelligence of a creature, the greater the capacity for learning becomes. Why would no correlation exist?
AI systems reason, make inferences, learn, re-program themselves, and deal with situations unanticipated by their programmers.
If this is truly the case, I would certainly lend more credence to the supposition, however, I would need to see some specifics before I impute any value to it. Tell me more about this in more detail if you don't mind.
They generally lack the ability to use common-sense reasoning, and are intelligent only within constrained domains. So, while we have nothing approaching the sci-fi versions of AI, neither is there any evidence of some sort of hard limit as to what mental abilities might be replicated artificially.
"Common sense" seems to be a more ambiguous term than intelligence does. What constitutes common sense and how or why are modern supercomputers incapable of understanding it? I would say that common sense is least form of intelligence. Case in point, its common to all or most, whereas, great mathematicians or great musical composers are a diamond in the rough-- meaning, they are more highly intelligent than the average person. That lends more acceptance to the fact that the computer is designed, not that it learns as it goes along.
Let me try to make this point more clearly. Either "intelligence" is a property that can be ascertained to exist in an entity, or it isn't.
Isn't that the way it is with everything? That's the law of non-contradiction which applies to virtually every facet of life. Either God exists or He doesn't. Either you are in Africa or South America. Either the answer to the formula is 5 or it isn't.
If it isn't, then ID is a big waste of time. If it is, then entity X is either intelligent or it is not.
ID may be a big waste of time just as anything else. But there is only one way to find out. Phrenology had to be tried to see if it indeed offered anything to science. Under scrutiny it didn't pan out, but only after some considerable testing.
If X is intelligent, then it makes no difference how X became intelligent.
That's silly. The "how" and "why" answer questions. If X is a Tsunami, are you saying that how X became a Tsunami is unimportant? Understanding how and why are obviously important. Indeed, that's what the entire premise of science is based on. The theory of evolution, from start to finish, is based on the how and why, otherwise the theory wouldn't exist.
Indeed, when ID critics ask IDists "Who designed the designer?", ID's response is that the ultimate cause of the Designer's intelligence is irrelevant!
Sure, you could make that argument. There could be a Designer for a Designer who designed the Designer's Designer. It could trek on for all eternity for all we know, but it doesn't speak anything to us.
Therefore, when we look at a computer that does intelligent things, according to ID, we can decide that it is intelligent per se, without considering its ultimate cause.
Sure, of course you can determine if something is intelligent without knowing the designer. Indeed, if I found a car in the middle of the ocean, I wouldn't need to know who the manufacturer is in order to deduce that it was intelligently designed.
Given we've established that computers are bona-fide intelligent agents
How can you establish if computers are bona-fide intelligent agents when you don't even have a working definition for what intelligence is?
it is irrelevant that they were designed by humans.
Irrelevant? Then let no company display their company logo if it is unimportant and remove your name from the peer reviewed dissertations.
If you deny computers are intelligent simply because they were designed, you must also deny that human beings are intelligent if you also believe that we were designed.
I've never made any such allusions. The fact that it was designed doesn't exclude it from intelligence. I'm simply questioning whether or not computers can be deemed as intelligent at all.
Now, what does this prove? It proves that processes that act strictly according to deterministic natural law, combined with random chance, can be intelligent. Both AI computers, and evolutionary processes, act strictly in accordance with deterministic law plus chance.
What? But you created it! Chance? Deterministic law? You made it the way it is. Chance plays no part in it. The computer didn't create itself, nor does it plot out its own destiny. It was assigned to them by the programmer. If you didn't come along, it wouldn't be here right now.
The fact that chance plus necessity can yield intelligent behavior does not prove that we were not intelligently designed.
What "necessity" faces computers? What will happen to the computer if no one updates the information and never changes the design? What consequence exists for the computer itself? If no consequence exists, then there is no necessity for it either. The only one facing necessity is the programmer who has to keep up with technology, otherwise his model will become obsolete and its marketability will fade. The computer faces no such necessity which makes the biological analogy useless.
It does, however, render the arguments of ID moot. ID claims that something that does not act according to purely natural chance and necessity is required to build the complex machinery of life, but since we have observable proof that nothing but chance and necessity is required for intelligence, ID's claim is unwarranted.
If someone has only half of the components for a computer, the entire computer will lose all functionality because all the components must exist simultaneously and in operable form, otherwise the entire mechanism fails. Necessity and chance literally play no part in it. That's just a hopeful monster.
I have not equated "intelligence" with anything.
That's part of the problem. You are being vague in one instance and then expect us to use common sense in another. You've yet to pin down what intelligence even constitutes, yet at the same time, want us to believe that computers are intelligent. That makes no sense.
I have said there is no scientifically useful definition for the word.
Then you can't scientifically explain that Artificial Intelligence is intelligent if the word "intelligence" is an ambiguous term. That's putting the cart before the horse. The logic doesn't follow.
This is a very naive view of AI. No, calculations are most certainly not "stored in an internal bank" (think about it - there are an infinite number of calculations that can be done in simple arithmetic).
Its naivete to assume that computers have a CPU where it makes those calculations? Is a calculator intelligent?
Rather, AI computers are "born" with some knowledge and abilities, and given additional knowledge by their programmers and by learning.
Right, "given" is the operative word. Chance plays no part in it.
As my colleague Drew McDermott memorably quipped, "Saying Deep Blue doesn't really think about chess is like saying an airplane doesn't really fly because it doesn't flap its wings."
That is a clever quip. I like it a lot, however, this proverb ultimately fails when juxtaposed with a little reason. If we were to assign specific qualitative properties to what "flying" actually is, then we could know what it means to fly. If flapping wings was apart of the criteria for true flying, then airplanes never once flew and still don't fly. However, if the criteria for flying is anything that can battle gravity for periods of time, then the airplane certainly does fly. Therefore, the definition of flying becomes just as important than the action. For you, "thinking" is about as cryptic as it could get. And because of your refusal to define intelligence and thinking, you're leaving it open to anything being considered capable of "thinking." Without a criterion, its as ambiguous as saying that rocks think.
AI has a very clear operational defintion of "intelligence": It is whatever people might generally tend to call intelligence.
Now you are flatly contradicting yourself. How can the definition be clear when you and I disagree on what intelligence is? How can we simply leave it to such relativity if you want your statement to have any meaning?
Again, please remember I am talking about scientific explanation, not causual conversation, philosophy, or theology. There is no scientific way to detect capriciousness.
If there is no scientific way to detect capriciousness, then there is no way to detect intelligence. If that's the case, then you can never make pronouncements on what is arbitrary and what is purposeful, either in nature or with software. You would render your own argument null and void.
If you choose "teleology" as your definition of intelligence, that's fine. As Norbert Wiener, the father of cybernetics, defined "teleology" about fifty years ago, it is the property of a system to work towards a goal.
I object to Weiner's definition, though some definition is better than none. His definition is too broad.
I of course do not believe that science can capture all truth (look at my signature line!).
Duly noted, good sir.
I actually happen to believe that our conscious awareness is something that transcends mechanism. In another context, I would happily discuss the merits of dualism, how quantum mechanics may or may not provide a way to understand free will, and so on.
I'd be very interested in hearing more about this. I don't know if that would take us off-topic, but if it did, we could always open a new thread. I'm sure it would generate a lot of attention because it sounds unique to the forum.
My honest answer is that there is no way to objectively describe what the similarity is between all "intelligent" creatures. Deciding whether or not some critter is intelligent or not is simply an exercise in definition, rather than discovery.
That's fine. I agree that if we leave it open to ambiguity, ambiguity so we shall receive. I'm objecting that in one instance you tell us to use common sense to see things that are obviously intelligent, but in the next, you remain vague on what intelligence actually is. Until a clear definition is given, I dare say that we're going to go around in circles until one or both of us loses interest.
I have told you exactly what I think "intelligence" is: A subjective, informal way we refer to some sorts of behaviors. Like "athletic" is a loose way to describe some sorts of physical abilities, or "beautiful" is a loose way to describe some sorts of physical features. None of these words are scientifically useful.
This is the most candid you have been, IMO. But science can and would give some kind of operational definition. The one thing you said that really caught my eye and that I believe will solidify your argument is whether or not the computer can learn and enact a will not programmed by the designer. That, to me, would be a clear case of legitimate cognizance.
Looks like I made it all the way through without catastrophe.
I think my orignal was better though.

"A man can no more diminish God's glory by refusing to worship Him than a lunatic can put out the sun by scribbling the word, 'darkness' on the walls of his cell." -C.S. Lewis

This message is a reply to:
 Message 35 by aiguy, posted 01-01-2007 3:06 PM aiguy has replied

Replies to this message:
 Message 42 by aiguy, posted 01-02-2007 3:46 PM Hyroglyphx has replied

  
TheMystic
Inactive Member


Message 41 of 71 (373726)
01-02-2007 2:13 PM
Reply to: Message 39 by aiguy
01-02-2007 1:57 PM


My objections are a whole lot simpler than what you list: I don't think evolution happened, not the 'molecules-to-man' kind of thing I presume you're talking about. So, have you got another example? The scientific method says we should be able to reproduce the results as often as possible, right?
Objection: But evolution can not create complex form and function (CFF)!
Response: Even if the currently understood mechanisms of evolutionary theory fail to account for all biological phenomena, we know that pure chance and necessity can indeed create CFF, since computers do it all the time.
I don't follow you here - what do you mean by computers creating CFF by pure chance and necessity? Oh, I bet you the evolutionary computer. This is circular reasoning, no?
Objection: But computers can only create CFF because they were designed by people!
Response: Once they are created, computers can create CFF. And they create CFF all by themselves, without guidance, by operating according to pure chance and necessity.
Well, no, I guess you do mean literal computers. I don't follow you - bits and bytes are my line of work, so I don't know what you mean by pure chance - other than using Johnson noise in cryptology or something, a computer is a deterministic beast (sometimes you doubt that, but sure enough you left a semicolon out somewhere).
And I don't know what you mean by 'chance and necessity' I guess. I think most people mean noise when they talk of chance, i.e. wideband excitation, but I don't know what you mean by necessity. Are you referring to the 'laws of nature' being deterministic?

This message is a reply to:
 Message 39 by aiguy, posted 01-02-2007 1:57 PM aiguy has replied

Replies to this message:
 Message 43 by aiguy, posted 01-02-2007 4:59 PM TheMystic has replied

  
aiguy
Inactive Member


Message 42 of 71 (373745)
01-02-2007 3:46 PM
Reply to: Message 40 by Hyroglyphx
01-02-2007 2:12 PM


Re: Computer models
Hi Nemesis -
Nemesis writes:
I started to respond to this post yesterday and was nearly finished. I got up to use the restroom and came back to a blank page. Apparently my daughter hit the back button. Yeah, that's frustrating. Hopefully I can get through the whole post this time.
I've learned a few things after programming for thirty years: Save, backup, save, backup, save, backup, and then backup some more.
Nemesis writes:
If there isn't a classical approach, then how is it that we humans generally agree on the same descriptions concerning intelligence?
Oh, we really do not agree at all! If you ask ten people for a definition of intelligence, you will get more than ten different answers.
Nemesis writes:
Just because science cannot, as of yet, understand what the will is does not negate the existence or that it is somehow inconsequential to intelligence.
Right, but it does mean we have no scientific way to talk about it.
Nemesis writes:
Would you agree that only intelligent beings are capable of asserting their will?
I think we have no idea. By "will", I presume you mean "contra-causal free will" or "libertarian free will". That is, something with "will" can break the known laws of physics and make something happen that has no antecedent physical cause. Well, there is no evidence that anything of the sort ever happens.
Nemesis writes:
I only ask because it seems that computers perform the tasks assigned to them and only those tasks.
No, that isn't true. Computer systems do things that their programmers have never anticipated. If a computer system incorporates randomness, it can come up with all sorts of new ideas, and select the ones it wishes to act upon by evaluating the new ideas according to fixed, deterministic processes. (sound familiar?)
Nemesis writes:
They don't exhibit any kind of freewill, whereas, natural beings enact their will at their discretion. I suppose one could make the argument that they too could only be performing the functions assigned to them. For face value I could agree with that. But even if this is so, then natural beings are still far more complex than even supercomputers. But even more than that, it says that those creatures were indeed designed with specifications.
Whew... you've gone through a number of different arguments here. First, think about this contra-causal free will thing, and hopefully you'll see that we can't determine if anything is capable of this or not, but we have no evidence that it does, and all of our understanding of the world so far indicates that nothing breaks the laws of physics.
Second, I don't think that the fact that computers are not yet as complex as living things is relevant. Why would it be? In any case, computers are as complex as some living things already, and they are getting more complex very rapidly.
Third, once a computer exists, it can generate novel, complex functional structures all by itself, purely by natural processes of chance and necessity. Therefore, we know that physical processes in nature, once they exist, can do the same thing. You may wish to argue that both computers and nature itself must have been created by something intelligent - that's fine with me, I won't argue that point.
Nemesis writes:
AIGUY: I would also exclude "emotion", since while humans seem to use emotion in their thinking, not all intelligent agents necessarily would.
NEMESIS: But this is the very thing that presents a problem for the secularist argument. For a secularist, biology is the only acceptable answer to the question, which invariably means that it is the only thing that offer any answers. An evolutionist will note that a sense of morality or the arts must have derived through natural means. If emotion is separate from intellect, then what inexplicable reason made it so? Where does emotion come from, and is it truly exclusive from intelligence?
I'm not sure what question you mean when you say "biology is the only acceptable answer to the question". I won't defend evolutionary psychological explanations for morality or aesthetics, even though I find them plausible; I'm willing to say we don't have explanations for these things. But this has no bearing on this point: Why would every intelligent thing have to have human-like emotions?
Think about a complete psychopath - a defective human who has no emotions. Isn't it possible that this person could still learn calculus and physics and build watches. Or think about another type of animal that does things you think are intelligent... are you sure that animal has human-like emotions?
Nemesis writes:
Learning is inconsequential to intelligence? Surely this is incorrect. The greater the intelligence of a creature, the greater the capacity for learning becomes. Why would no correlation exist?
Again, these questions are simply matters of definition. If you'd like to define intelligence such that learning is a necessary component, then you have just eliminated an omniscient creator as an intelligent thing (since an omniscient being already knows everything, so there is nothing to learn).
Nemesis writes:
AIGUY: AI systems reason, make inferences, learn, re-program themselves, and deal with situations unanticipated by their programmers.
NEMESIS: If this is truly the case, I would certainly lend more credence to the supposition, however, I would need to see some specifics before I impute any value to it. Tell me more about this in more detail if you don't mind.
Which part do you doubt? Reasoning and inferences - standard AI techniques. Learning - a whole discipline called machine learning. Self-programming - artificial neural networks are all self programming, and there are other techniques as well (e.g. Self-programming machines (II): Network of self-programming machines driving an Ashby homeostat - PubMed). Dealing with novel situations - think about the DARPA autonomous vehicle challenge; all the robot vehicles had to deal with all sorts of unanticipated situations.
Nemesis writes:
"Common sense" seems to be a more ambiguous term than intelligence does. What constitutes common sense and how or why are modern supercomputers incapable of understanding it?
Yes, an ambiguous term. We mean all of the zillions of things that humans beings are born knowing, or learn simply by existing in the world.
Nemesis writes:
I would say that common sense is least form of intelligence. Case in point, its common to all or most, whereas, great mathematicians or great musical composers are a diamond in the rough-- meaning, they are more highly intelligent than the average person. That lends more acceptance to the fact that the computer is designed, not that it learns as it goes along.
It turns out that the hardest things for humans to learn - advanced mathematics and theorem proving, championship chess playing, medical and machine diagnosis, etc - these things are relatively easy for computers to learn. Common sense, however - if you turn a glass of water upside-down over your lap, you will get wet - that is very hard for computers. Read about it here: Cyc | The Next Generation of Enterprise AI.
Nemesis writes:
AIGUY: Let me try to make this point more clearly. Either "intelligence" is a property that can be ascertained to exist in an entity, or it isn't.
NEMESIS: Isn't that the way it is with everything? That's the law of non-contradiction which applies to virtually every facet of life. Either God exists or He doesn't. Either you are in Africa or South America. Either the answer to the formula is 5 or it isn't.
What I said was either intelligence could be ascertained to exist or not. This is not the case for everything, no, since in order to ascertain if a property exists in some entity, that property must be coherently defined. I can't answer the question of God eists or He doesn't, for example, because I honestly don't know what you mean by God.
Nemesis writes:
ID may be a big waste of time just as anything else. But there is only one way to find out. Phrenology had to be tried to see if it indeed offered anything to science. Under scrutiny it didn't pan out, but only after some considerable testing.
No, not everything must be - or can be - tried to see if it offers anything to science. Take my theory of gunderplitzen, for example. I believe that people are intelligent because they contain gunderplitzen. Do you think we ought to give this theory a scientific chance in order to see if it's true? (Don't ask me what gunderplitzen is - I'm not really sure yet).
Nemesis writes:
AIGUY: If X is intelligent, then it makes no difference how X became intelligent.
NEMESIS: That's silly. The "how" and "why" answer questions. If X is a Tsunami, are you saying that how X became a Tsunami is unimportant?
Yes - of course. If X is a tsunami, then X is a tsunami no matter how X became a tsunami. If X is intelligent, then X is intelligent no matter how X became intelligent.
Nemesis writes:
Understanding how and why are obviously important. Indeed, that's what the entire premise of science is based on. The theory of evolution, from start to finish, is based on the how and why, otherwise the theory wouldn't exist.
I'm not saying that we would not like to know things. I'm saying that the way X became intelligent does not alter the fact that X is intelligent.
Nemesis writes:
AIGUY: Indeed, when ID critics ask IDists "Who designed the designer?", ID's response is that the ultimate cause of the Designer's intelligence is irrelevant!
NEMESIS: Sure, you could make that argument. There could be a Designer for a Designer who designed the Designer's Designer. It could trek on for all eternity for all we know, but it doesn't speak anything to us.
My point here is that ID itself holds that intelligence can be "detected" without regard to the source of the intelligence.
Nemesis writes:
Sure, of course you can determine if something is intelligent without knowing the designer. Indeed, if I found a car in the middle of the ocean, I wouldn't need to know who the manufacturer is in order to deduce that it was intelligently designed.
No, I meant that you can decide something is intelligent without any consideration of how that thing became intelligent. Therefore, if a computer does something you think is intelligent, the fact that a human designed the computer does not disqualify the computer from being a bona-fide intelligence. Likewise, if a human does something you think is intelligent, the fact that you think the human was designed by the Designer does not disqualify the human from being a bona-fide intelligence.
Nemesis writes:
How can you establish if computers are bona-fide intelligent agents when you don't even have a working definition for what intelligence is?
If you think the term is meaningless, that's great, we're done. A scientific theory can't offer meaningless terms as explanations for things. I have already said many times now what I think the word means (a loose way of describing various things that strike us subjectively as being descriptive of something's mental abilities, like "athleticsm" describes some physical abilities or "beauty" describes some physical features).
Nemesis writes:
AIGUY: it is irrelevant that they were designed by humans.
NEMESIS: Irrelevant? Then let no company display their company logo if it is unimportant and remove your name from the peer reviewed dissertations.
Come on now. The fact that computers are designed by humans is irrelevant to the question of whether or not computers are intelligent, just like the fact that you think humans were designed by a Designer is irrelevant to the question of whether or not humans are intelligent.
Nemesis writes:
I've never made any such allusions. The fact that it was designed doesn't exclude it from intelligence. I'm simply questioning whether or not computers can be deemed as intelligent at all.
Computers can reason, perform inferences, create novel designs of irreducibly complex machinery. They can induce rules from examples, deduce facts from other facts, diagnose strange medical disorders, detect credit card fraud, fix a deep space probe all by themselves, drive a car, and beat the pants off of either one of us in a chess game...
Nemesis writes:
AIGUY: Now, what does this prove? It proves that processes that act strictly according to deterministic natural law, combined with random chance, can be intelligent. Both AI computers, and evolutionary processes, act strictly in accordance with deterministic law plus chance.
NEMESIS: What? But you created it! Chance? Deterministic law? You made it the way it is. Chance plays no part in it. The computer didn't create itself, nor does it plot out its own destiny. It was assigned to them by the programmer. If you didn't come along, it wouldn't be here right now.
Please read my response to Mystic, above. Once the computer exists, it can do all these intelligent things purely by processes of chance and necessity. Once nature exists, then, it can do all these intelligent things purely by processes of chance and necessity. If you would like to argue that both computers and nature itself had an intelligent designer, I won't argue that. However, you can't argue that nature itself cannot do intelligent things (like create complex form and function of biology) all by itself, by processes of pure chance and necessity.
Nemesis writes:
What "necessity" faces computers?
I mean the deterministic physical laws (electricity, etc) according to which computers operate. The "chance" of computer systems comes from random (or pseudo-random) input.
Nemesis writes:
What will happen to the computer if no one updates the information and never changes the design? What consequence exists for the computer itself? If no consequence exists, then there is no necessity for it either. The only one facing necessity is the programmer who has to keep up with technology, otherwise his model will become obsolete and its marketability will fade. The computer faces no such necessity which makes the biological analogy useless.
The analogy is between the computer system and natural processes - like evolution. Both operate according to pure chance and necessity, i.e. physical law plus randomness.
Nemesis writes:
AIGUY: I have not equated "intelligence" with anything.
NEMESIS: That's part of the problem. You are being vague in one instance and then expect us to use common sense in another. You've yet to pin down what intelligence even constitutes, yet at the same time, want us to believe that computers are intelligent. That makes no sense.
My point is that ID offers "intelligence" as the explanation for biology, but "intelligence" has no scientifically useful definition. That isn't my fault.
Nemesis writes:
AIGUY: I have said there is no scientifically useful definition for the word.
NEMESIS: Then you can't scientifically explain that Artificial Intelligence is intelligent if the word "intelligence" is an ambiguous term. That's putting the cart before the horse. The logic doesn't follow.
AI defines intelligence simply as "that which, if a person did it, we would tend to call it intelligent". That's all.
Nemesis writes:
Its naivete to assume that computers have a CPU where it makes those calculations?
Most computers have CPUs, yes. But the calculations are not "stored" - the computer figures them out when it needs to.
Nemesis writes:
Is a calculator intelligent?
There is no scientific fact of the matter. It is purely a matter of definition, and not discovery.
Nemesis writes:
AIGUY: Rather, AI computers are "born" with some knowledge and abilities, and given additional knowledge by their programmers and by learning.
NEMESIS: Right, "given" is the operative word. Chance plays no part in it.
Just like humans are "given" additional knowledge by parents and teachers. And chance does play a part in it when computer systems employ randomness.
Nemesis writes:
AIGUY: As my colleague Drew McDermott memorably quipped, "Saying Deep Blue doesn't really think about chess is like saying an airplane doesn't really fly because it doesn't flap its wings."
NEMESIS: That is a clever quip. I like it a lot, however, this proverb ultimately fails when juxtaposed with a little reason. If we were to assign specific qualitative properties to what "flying" actually is, then we could know what it means to fly. If flapping wings was apart of the criteria for true flying, then airplanes never once flew and still don't fly. However, if the criteria for flying is anything that can battle gravity for periods of time, then the airplane certainly does fly. Therefore, the definition of flying becomes just as important than the action.
Just so. The plane does what it does, and it doesn't matter if you call it "flying" or not. The computer does what it does, and it doesn't matter if you call it "thinking", or "intelligent", or not.
Nemesis writes:
For you, "thinking" is about as cryptic as it could get. And because of your refusal to define intelligence and thinking, you're leaving it open to anything being considered capable of "thinking." Without a criterion, its as ambiguous as saying that rocks think.
When we define "flying", we can come up with a definition that independent researchers can measure and agree on. "Flying" can be given an unambiguous, clear, empirically grounded meaning: It can mean "move above the surface of the Earth supported only by air pressure", for example. The same is not true of intelligence - there is no unambiguous, emprically grounded meaning for that term that independent researchers can test and replicate.
Nemesis writes:
AIGUY: AI has a very clear operational defintion of "intelligence": It is whatever people might generally tend to call intelligence.
NEMESIS: Now you are flatly contradicting yourself. How can the definition be clear when you and I disagree on what intelligence is? How can we simply leave it to such relativity if you want your statement to have any meaning?
You are arguing that "intelligence" has no clear meaning that we can agree on. I agree with you!!! So how can you have a scientific theory that offers "intelligence" as an explanation???
Nemesis writes:
If there is no scientific way to detect capriciousness, then there is no way to detect intelligence.
Agreed!!!
Nemesis writes:
If that's the case, then you can never make pronouncements on what is arbitrary and what is purposeful, either in nature or with software. You would render your own argument null and void.
My argument is that "intelligence" cannot be used to explain things. Nothing is explained by saying "it is intelligent".
Q: How does a termite colony build those complicated tower complexes?
A: Because termite colonies are intelligent.
Q: How do you know termite colonies are intelligent?
A: Because that is what we tend to call things that build complicated structures.
Nemesis writes:
I object to Weiner's definition, though some definition is better than none. His definition is too broad.
His definition brought goals and purpose into the realm of science. What it leaves out - if anything - is what can't be scientifically analyzed.
Nemesis writes:
AIGUY: I actually happen to believe that our conscious awareness is something that transcends mechanism. In another context, I would happily discuss the merits of dualism, how quantum mechanics may or may not provide a way to understand free will, and so on.
NEMESIS: I'd be very interested in hearing more about this. I don't know if that would take us off-topic, but if it did, we could always open a new thread. I'm sure it would generate a lot of attention because it sounds unique to the forum.
I have been fascinated by the mind/body problem since I was a little boy, and I've been studying philosophy of mind for more than three decades. Yes, at some point I'll start a thread in another forum, but I can only manage to participate in one thread at a time (I spend too much time as it is on this!!)
Nemesis writes:
AIGUY: That's fine. I agree that if we leave it open to ambiguity, ambiguity so we shall receive. I'm objecting that in one instance you tell us to use common sense to see things that are obviously intelligent, but in the next, you remain vague on what intelligence actually is. Until a clear definition is given, I dare say that we're going to go around in circles until one or both of us loses interest.
If you agree that we have no useful scientific definition of intelligence, I will be quite satisfied: My overriding point is that there is no scientific warrant for inferring "intelligence" as the explanation for biological complexity.
Nemesis writes:
AIGUY: I have told you exactly what I think "intelligence" is: A subjective, informal way we refer to some sorts of behaviors. Like "athletic" is a loose way to describe some sorts of physical abilities, or "beautiful" is a loose way to describe some sorts of physical features. None of these words are scientifically useful.
NEMESIS: This is the most candid you have been, IMO.
I always aim for complete disclosure... it's just that we have to talk for a while before we can figure out what each other thinks.
Nemesis writes:
But science can and would give some kind of operational definition. The one thing you said that really caught my eye and that I believe will solidify your argument is whether or not the computer can learn and enact a will not programmed by the designer. That, to me, would be a clear case of legitimate cognizance.
Here's an excellent, free, introductory text on machine learning online: Error 404 Not Found | Stanford Artificial Intelligence Laboratory
Nemesis writes:
Looks like I made it all the way through without catastrophe.
I think my orignal was better though.
Save, backup, then backup some more...

Science is not simply reason - it is much less than that. It is reason constrained by empiricism.

This message is a reply to:
 Message 40 by Hyroglyphx, posted 01-02-2007 2:12 PM Hyroglyphx has replied

Replies to this message:
 Message 57 by Hyroglyphx, posted 01-03-2007 6:38 PM aiguy has replied

  
aiguy
Inactive Member


Message 43 of 71 (373769)
01-02-2007 4:59 PM
Reply to: Message 41 by TheMystic
01-02-2007 2:13 PM


Hi Mystic,
Mystic writes:
I don't follow you here - what do you mean by computers creating CFF by pure chance and necessity? Oh, I bet you the evolutionary computer. This is circular reasoning, no?
Genetic algorithms are one type of AI program that can create CFF by pure chance and necessity. By "chance" I mean randomness (in computer systems, randomness is accessed either via pseudo-random generators or by physical random process transducers). By "necessity" I mean according to the laws of physics.
Mystic writes:
AIGUY: Objection: But computers can only create CFF because they were designed by people!
Response: Once they are created, computers can create CFF. And they create CFF all by themselves, without guidance, by operating according to pure chance and necessity.
MYSTIC: Well, no, I guess you do mean literal computers.
Once created, both nature and computer systems can create CFF by means of pure chance and necessity.
Mystic writes:
I don't follow you - bits and bytes are my line of work, so I don't know what you mean by pure chance - other than using Johnson noise in cryptology or something, a computer is a deterministic beast (sometimes you doubt that, but sure enough you left a semicolon out somewhere).
So you already know that some computer programs (like GAs) use random input, and that input can be pseudo-random or "real" randomness (like radio static or radioactive decay intervals).
Mystic writes:
And I don't know what you mean by 'chance and necessity' I guess. I think most people mean noise when they talk of chance, i.e. wideband excitation, but I don't know what you mean by necessity. Are you referring to the 'laws of nature' being deterministic?
Yes, and yes. Once again:
1) Once computers comes to exist, they can do things that ID folks call "intelligent" - like creating irreducibly complex machines. The way computers do this is by operating according to nothing but randomness + deterministic physical law, because that is how computers work.
2) Once nature comes to exist, it can do things that ID folks call "intelligent" - like creating irreducibly complex biological structures. The way nature does this is by operating according to nothing but randomness + deterministic physical law, because that is how nature works (as far as we know).
If you'd like to say that both computers and nature itself were designed by something intelligent, go right ahead - I won't argue that point.
Can you see a flaw in that argument?

Science is not simply reason - it is much less than that. It is reason constrained by empiricism.

This message is a reply to:
 Message 41 by TheMystic, posted 01-02-2007 2:13 PM TheMystic has replied

Replies to this message:
 Message 44 by TheMystic, posted 01-03-2007 10:45 AM aiguy has replied

  
TheMystic
Inactive Member


Message 44 of 71 (373967)
01-03-2007 10:45 AM
Reply to: Message 43 by aiguy
01-02-2007 4:59 PM


1) Once computers comes to exist, they can do things that ID folks call "intelligent" - like creating irreducibly complex machines.
If nothing else you can observe that they do it for only a short time. Left to themselves computers will turn into rust buckets within a few decades at best. Nature, by contrast, is alleged to have steadily and without intervention increased in complexity for 10s (100s?) of millions of years. (Yes, I know about the 'evolution in bursts' idea).

This message is a reply to:
 Message 43 by aiguy, posted 01-02-2007 4:59 PM aiguy has replied

Replies to this message:
 Message 45 by aiguy, posted 01-03-2007 12:42 PM TheMystic has replied

  
aiguy
Inactive Member


Message 45 of 71 (374003)
01-03-2007 12:42 PM
Reply to: Message 44 by TheMystic
01-03-2007 10:45 AM


Hi Mystic,
aiguy writes:
1) Once computers comes to exist, they can do things that ID folks call "intelligent" - like creating irreducibly complex machines.
Mystic writes:
If nothing else you can observe that they do it for only a short time. Left to themselves computers will turn into rust buckets within a few decades at best. Nature, by contrast, is alleged to have steadily and without intervention increased in complexity for 10s (100s?) of millions of years. (Yes, I know about the 'evolution in bursts' idea).
I've presented a simple argument that undermines Intelligent Design Theory:
Computers can generate CSI
Computers operate according to natural processes
Therefore natural processes (chance and necessity) can generate CSI
Therefore there is no justification to posit anything else to account for biology
Your rebuttal is: Computers don't live very long.
Perhaps you'll be able to tie this into my argument somehow, but as far as I can tell, this is what one would call a red herring.

Science is not simply reason - it is much less than that. It is reason constrained by empiricism.

This message is a reply to:
 Message 44 by TheMystic, posted 01-03-2007 10:45 AM TheMystic has replied

Replies to this message:
 Message 46 by TheMystic, posted 01-03-2007 12:58 PM aiguy has replied

  
Newer Topic | Older Topic
Jump to:


Copyright 2001-2023 by EvC Forum, All Rights Reserved

™ Version 4.2
Innovative software from Qwixotic © 2024