rwojcik@bcsaic.UUCP (Rick Wojcik) (04/07/88)
In article <1134@its63b.ed.ac.uk> gvw@its63b.ed.ac.uk (G Wilson) writes: >[re: my reference to natural language programs] >Errmmm...show me *any* program which can do these things? To date, >AI has been successful in these areas only when used in toy domains. > NLI's Datatalker, translation programs marketed by Logos, ALPs, WCC, & other companies, LUNAR, the LIFER programs, CLOUT, Q&A, ASK, INTELLECT, etc. There are plenty. All have flaws. Some are more "toys" than others. Some are more commercially successful than others. (The goal of machine translation, at present, is to increase the efficiency of translators--not to produced polished translations.) >... Does anyone think AI would be as prominent >as it is today without (a) the unrealistic expectations of Star Wars, >and (b) America's initial nervousness about the Japanese Fifth Generation >project? > I do. The Japanese are overly optimistic. But they have shown greater persistence of vision than Americans in many commercial areas. Maybe they are attracted by the enormous potential of AI. While it is true that Star Wars needs AI, AI doesn't need Star Wars. It is difficult to think of a scientific project that wouldn't benefit by computers that behave more intelligently. >Manifest destiny?? A century ago, one could have justified >continued research in phrenology by its popularity. Judge science >by its results, not its fashionability. > Right. And in the early 1960's a lot of people believed that we couldn't land people on the moon. When Sputnik I was launched my 5th grade teacher told the class that they would never orbit a man around the earth. I don't know if phrenology ever had a respectable following in the scientific community. AI does, and we ought to pursue it whether it is popular or not. >I think AI can be summed up by Terry Winograd's defection. His >SHRDLU program is still quoted in *every* AI textbook (at least all >the ones I've seen), but he is no longer a believer in the AI >research programme (see "Understanding Computers and Cognition", >by Winograd and Flores). Weisenbaum's defection is even better known, and his Eliza program is cited (but not quoted :-) in every AI textbook too. Winograd took us a quantum leap beyond Weisenbaum. Let's hope that there will be people to take us a quantum leap beyond Winograd. But if our generation lacks the will to tackle the problems, you can be sure that the problems will wait around for some other generation. They won't get solved by pessimists. Henry Ford had a good way of putting it: "If you believe you can, or if you believe you can't, you're right." -- Rick Wojcik csnet: rwojcik@boeing.com uucp: {uw-june uw-beaver!ssc-vax}!bcsaic!rwojcik address: P.O. Box 24346, MS 7L-64, Seattle, WA 98124-0346 phone: 206-865-3844
doug@isishq.UUCP (Doug Thompson) (04/12/88)
rwojcik@bcsaic.UUCP (Rick Wojcik) writes:
RJ> Weisenbaum's defection is even better known, and his Eliza program
RJ> is
RJ> cited (but not quoted :-) in every AI textbook too. Winograd
RJ> took us a
RJ> quantum leap beyond Weisenbaum. Let's hope that there will be
RJ> people to take
RJ> us a quantum leap beyond Winograd. But if our generation lacks
RJ> the will
RJ> to tackle the problems, you can be sure that the problems will
RJ> wait
RJ> around for some other generation. They won't get solved by pessimists.
Weisenbaum's problem with AI does not fairly translate into "pessimism"
or a lack of "will". Sure, he points out that many forecasts of
breakthroughts by AI types just haven't been realized. His complaint, in
the book "Computer Power and Human Reason" seems much more about the
mentality of some AI workers, and the way problems are defined.
He disputes, for instance, the populist scientific view of man, that
homo-sapiens is just a complex machine (behaviourism). Machines depend
on "effective procedures" -- logical if-then and cause-effect
relationships. Artificial intelligence, so long as it is based on the
kind of machines we know about today, must also be based on "effective
procedures". Human decision-making, according to Weisenbaum, makes use
of "effective-procedures" at times, but is not confined to it. Human
decision-making usually involves two or more people talking about a
problem and coming to understand its meaning to them, as distinct
persons, in an historical situation. This suggests avenues for solution.
The way we talk to each other, the way we understand meaning -- this has
much to do with the experience of being a human being, and being treated
like a human being by other human beings.
How are you going to mechanize that? Either the machine will understood
pre-defined meanings "programmed" into it, or it will develop its own
meanings based on its own experience. The latter is far beyond current
capabilities, and the former is -- well -- relatively trivial. All you
end up with is a decision-making loop which is only successful it can
take account of *all* possible input and permutations. It is basically
no different than the rules of a welfare bureaucracy, for instance. The
raw data is the applicant, the applicant is examined according to
certain pre-defined criteria, and the bureaucracy decides to pay or not
pay. We all know this is unfair to some because people who don't really
need help get it, and some who really do need help don't. The real
humans applying for welfare don't always fit the pre-defined categories.
You could mechanize that process though, because it is based on clear
rules that are expressed as effective procedures. You might even call it
intelligence, but it is still not going to replace the human appeal
committee that can look at what the "machine" or the "bureaucracy"
decided and over-rule it when an exceptional case arises.
This is just one of the problems for which AI has, in Weisenbaum's
argument, no theoretical solution. People deal with new situations with
creativity, often through such things as empathy, based on their
experience of being human, and what it means to be human. Can you make a
machine think it is a human and think like a human?
Well, there are those who say you can -- there are those who say you
can't and Weisenbaum is saying you probably can't but you most certainly
*should not*.
Such a machine, if it worked, could end up imposing its creator's sense
of meaning, mostly frozen in time, on everyone subject to such a
machine. It could generate its own sense of meaning and run the world
according to what it -- not its creators or subjects -- thought was
important. Already there is lots of evidence that the limited
instruments we have today are doing this.
Ultimately the perfect AI machine would behave exactly as a human might,
and have all the capabilities that a human has. One seriously has to ask
why one would want to make a mechanical man for trillions of dollars
when we can get billions of flesh-and-blood men for pennies per hour?
I think we have here "optimism" based on a combination of ancient dreams
of "perfect slaves" and "supermen". Of course, we presume that we will
be able to control these machines once we have built them. That is the
romatic misconception. We build bureaucracies we cannot control,
institutions we cannot control, we have all written computer programs or
parts of programs that no one can control -- or properly understand.
I very much identify with Weisenbaum's basic question: "Why on earth
would we want such machines?" "What possible *good* could they do for
us?" The debate is partly technical, but mostly philosophic: grated for
the moment that you may eventually build such a thing (which is quite
doubtful) what would you use it for?
The answer, if you look about, is to make more effective military
weapons. Or at least this represents the majority of answers being found
today. Very little AI work is being directed toward reconciling human
differences, resolving world problems, feeding the poor, or bringing
justice to the oppressed. Very much AI work is being directed toward
increasing the capacity of some men, in possession of these instruments,
to control other men.
Since most AI advocacy is rooted in a behaviouristic understanding of
mankind, it is not surprising that instruments to modify behaviour
comprise most of what is being produced -- or researched. At best
though, we could "artificialize" only a very small and specific portion
of human "intelligence" by pursuing this path. Theologians define
idolatry as the worship of a sub-set of human attributes at the expense
of others, leaving an unbalanced, distorted result.
This is precisely Weisenbaum's complaint against AI. He doesn't say
there is not good to be achieved down this road, he does say that the
approach being taken in this culture is very unbalanced, and unhealthy
and therefore the net effect is negative.
His plea is not that we stop work on AI, but that we approach it in a
more balanced and wholistic way with a more civilized list of human
priorities such that the machines we create serve to benifit mankind,
and not make life more tenuous and intolerable.
At the moment, this is largely impossible for a wide variety of
political reasons. Weisenbaum was not prepared to work on AI devices to
enhance the kill power of military equipment, nor was he prepared to
work on machines to mechanize psychotherapy and remove the human doctor
from the treatment of humans. AI apologists generally are quite prepared
to work on such projects, and even hail them as great progress for the
human race. To Weisenbaum -- and me -- such things are Frankenstinian
obscenities which can only degrade human life.
RJ> Henry Ford had a good way of putting it: "If you believe you
RJ> can, or if
RJ> you believe you can't, you're right."
Well, I'm not gonna knock the power of "positive thinking" -- but Hitler
believed he could . . .
In addition to the question of "do you believe?" we must ask "in what do
you believe?" before deciding to help you or put a stop to you.
------------------------------------------------------------------------
Fido 1:221/162 -- 1:221/0 280 Phillip St.,
UUCP: !watmath!isishq!doug Unit B-3-11
Waterloo, Ontario
Bitnet: fido@water Canada N2L 3X1
Internet: doug@isishq.math.waterloo.edu (519) 746-5022
------------------------------------------------------------------------
---
* Origin: ISIS International H.Q. (II) (Opus 1:221/162)
SEEN-BY: 221/162
elg@killer.UUCP (Eric Green) (06/04/88)
in article <48.22A3B84F@isishq.UUCP>, doug@isishq.UUCP (Doug Thompson) says: > I wonder. I actually wonder if the human psyche (the subject of > psychology) can actually be dealt with "scientifically" or > mathematically at all. > Humans appear to be unpredictable. > Second, think about how you might scientifically or mathematically > analyse why John Doe is a republican and Jane Doe is a Democrat. Humans APPEAR to be unpredictable. However, psychiatrists today can to large extent predict how a certain person will react to certain situations, if they have a large enough base of knowledge about that person. E.G. the "stress tests" given to certain persons in critical positions to ascertain whether they will react correctly when an emergency situation arises. To generalize, there are a couple of basic assumptions needed to make AI a science instead of a religion: 1) The human mind consists of mechanism (program) and data (memories), 2) All human action is detirmined by the operation of the mind's mechanisms upon the mind's memories, 3) The computer can model the above. In particular, 1) neurologists have identified some of the mechanism by which the brain controls the body and stores memory (although they have not come anywhere close to understanding enough to help AI researchers much), and 2) Psychologists, via a rough understanding of some of the common basic mechanisms (e.g. the "pleasure principle"), and with a few of the patient's memories to apply those mechanisms to, have been fairly successful in uncovering some of the mechanism/memory juxtapositions underlying a particular action by a particular patient. So I have little doubt that the human mind IS succeptable to scientific analysis. As to whether it can be modelled (part 3), however, is still in doubt. For example, if we accept that a Z-80 is about as powerful as a neuron, it would still take over a million of them to approach the complexity of the human brain. Clearly, we are still quite a few years away from having the hardware horsepower necessary. > this into an AI model, and ask a machine to decide which is "best" or > which is "right", the democratic party or the republican party. Try to > replicate the human decision-making process at the ballot-box. Under the above model, your choice is dictated by your past experiences (memories) and by various mechanisms operating upon those memories (e.g. mechanism: pleasure/security. Memories: Political discussions, which policies were best, who was the better speaker, smarter, etc.). That is, your choice consists of neurons firing in a particular pattern. And, if somehow, microseconds before the vote, we had the sum total of your knowledge and experience available to us, we could predict exactly how you would vote. You could have voted in no other way besides that dictated to you by your knowledge and experience and the underlying mechanisms. > AI wants to build machines that can perform tasks or make decisions as > well as humans. I think though, that human reason and decision making is > not mechanical, it is a-mathematical, a-scientific and a-rational. It is > hinged to something else, human values (variable), passion (wholly > subjective), emotions (volatile) and sympathies (unpredictable). Science asserts the exact opposite: That everything can be explained, that everything can be modelled, and that the way to gain knowledge is to seek these explanations and models. Religion, on the other hand, is completely the opposite. Religion asserts that there are things that cannot be explained, that some things are beyond comprehension, and to seek explanation is to lose the faith. In other words, your statements imply that you are a member, unwittingly or no, of a religion whose basic tenate is that there is at least one thing in the universe (the human mind) which cannot be explained. Just as other religions have as their basic tenate that the one unexplainable thing is "God." > created and not instrinsically creative. My hunch is that human thought > is really dependent on dimensions of the universe which science (as we > currently understand it) is not yet capable of fathoming. I would be interested in hearing the reasoning behind your "hunch". If you have any evidence or experience that may be relevant to the AI community, perhaps we should hear of it? > How can you apply math or science to such things as Faith, Spiritual > sensibility, relgious experience, love or hatred? Oh boy. Want me to give you a long anthropological and/or sociological treatise upon the relation between faith in shamans, and survival in primitive societies? Love as an outgrowth of the nurturing instinct which assures survival of offspring? Hatred as an outgrowth of the warding instinct which protects the resources of the family group thus aiding survival? There has been much scientific research into the areas that you mention. In many of them, we do have some knowledge of the evolution of various social more's from ancient times to the present. Almost invariably, the root cause is some behavior which, at that time, had some positive survival value. An assertation that such behaviors cannot be scientifically explained is thus contradicted by the fact that we have sciences called History, Anthropology, Sociology, and Psychology. > Science can do very well with the natural world, but I suspect there is > a part of the human being which is strongly connected to a super-natural > reality which science has yet to get a grip on. Here is where I become certain that you are advocating a certain set of religious beliefs (commonly called "New Age", I believe). As I said before, the whole underpinning of science is that the world can be explained. That approach has yielded every scientific advance in the world today, from the wheat upon your table (a scientifically-bread hybrid, high yield), to the car you drive to work. There is no reason to expect that paridigm of reality to stop yielding results any time soon. On the other hand, the opposite paridigm (that there are things that cannot be explained, and to search is to lose your faith), has yielded no results at all besides innumerable books full of moral advice that everybody seems to ignore. To say that artificial intelligence cannot be attained because of magical or mystical properties of the human mind is thus of no relevance to scientists. Similiar statements have been made about other scientific endeavors, and have always proven false. The stumbling blocks to AI are much more mundane: the large bulk and limited computing power of today's computers, providing enough input devices to provide a suitable store of memories upon which to operate, etc. -- Eric Lee Green {cuae2,ihnp4}!killer!elg Snail Mail P.O. Box 92191 Lafayette, LA 70509 "Is a dream a lie if it don't come true, or is it something worse?"
doug@isishq.UUCP (Doug Thompson) (06/07/88)
>From: elg@killer.UUCP (Eric Green) > >Humans APPEAR to be unpredictable. However, psychiatrists today can to large >extent predict how a certain person will react to certain situations, >if they have a large enough base of knowledge about that person. E.G. >the "stress tests" given to certain persons in critical positions to ascertain >whether they will react correctly when an emergency situation arises. I'm not meaning to dispute that human behaviour displays a large degree of predictability. I am meaning to state that it also displays a large degree of *unpredictability*. I would also state that it is in the area of the unpredictability that most of the most interesting human behaviour occurs; for this is the province of creativity and genius. More on that in a moment. >To generalize, there are a couple of basic assumptions needed to make >AI a science instead of a religion: > >1) The human mind consists of mechanism (program) and data (memories), >2) All human action is detirmined by the operation of the mind's >mechanisms upon the mind's memories, >3) The computer can model the above. Agreed entirely. *IF* the human mind really is nothing more than a complex computer (mechanism) processing definable data, there is no reason to suppose that we cannot eventually build a machine that cannot do roughly the same sort of thing. >In particular, > 1) neurologists have identified some of the mechanism by which >the brain controls the body and stores memory (although they >have not come anywhere close to understanding enough to help AI researchers >much), That's not saying much. Few would argue that some human behaviour appears to be quite mechanical; i.e. a particiular stiumulus usually gets you a predictable result. That it hasn't helped AI much probably relates to the fact that what neurologists have discovered has next to nothing to do with thought and intelligence. > and > 2) Psychologists, via a rough understanding of some of the >common basic mechanisms (e.g. the "pleasure principle"), and with a few of >the patient's memories to apply those mechanisms to, have been fairly successful >in uncovering some of the mechanism/memory juxtapositions underlying >a particular action by a particular patient. Yep, I'm familiar with some very good work on this topic. Again though, you do not demonstrate that you can answer all the questions just by answering one or two of them. I can see where you might derive encouragement from this work, but it really doesn't *prove* that all thought, especially creative or compassionate thought, is merely mechanical. Do you really believe that you and I are nothing more than potent computers? >So I have little doubt that the human mind IS succeptable to >scientific analysis. As to whether it can be modelled (part 3), however, And therein I think you have expressed the problem clearly. You have little doubt. It was Descartes who said that the only thing really certain to him was his own capacity to doubt. It is critical to knowing anything. I'd refer you to the history of scientific revolutions, in which we find *doubt* is the primary engine of creativity and new discovery. >is still in doubt. For example, if we accept that a Z-80 is about as powerful >as a neuron, it would still take over a million of them to approach the complexity >of the human brain. Clearly, we are still quite a few years away from >having the hardware horsepower necessary. Ah! Classic, but it doesn't wash. Yeah, we don't have the CPU cycles to replicate everything that my brain does, but surely we have enough horsepower in super-computers to take just one highly intelligent process that I can accomplish in seconds and get the computer to do it in a year. We have the computing power for that. But we cannot do it. We can teach computers to play an excellent game of chess, but we know that the computer's approach and my approach to the chess board are totally different. >Under the above model, your choice is dictated by your past experiences >(memories) and by various mechanisms operating upon those memories >(e.g. mechanism: pleasure/security. Memories: Political discussions, >which policies were best, who was the better speaker, smarter, etc.). > >That is, your choice consists of neurons firing in a particular >pattern. And, if somehow, microseconds before the vote, we had the sum total >of your knowledge and experience available to us, we could predict exactly >how you would vote. You could have voted in no other way besides that >dictated to you by your knowledge and experience and the underlying >mechanisms. Which reminds me of a story. When I was about seven, I was wrestling with the problem of determinism. (It was in a theological context at the time) I stood at the head of the stairs and said to myself, "If God *really* knows everything, past present and future, God knows whether I will go down the stairs now, or later . . . but I don't know." It's a similar problem to the one you present, except the "determinism" in your position is mechanistic rather than theological. Still, it comes out to the same thing, that certain knowledge of exactly what I will do next does exist, the only difference is that you admit you can't actually attain that knowledge (yet). My difficulty with religious determinism and mechanistic determinism is pretty much the same. Subjectively I experience myself as a decision-makaing entity in possession of free will. I can choose. You and the Calvinists would both argue that there is really no choice at all, that the action I will choose tomorrow is completely knowable now. The Calvinists more simply assert that it is knowable (at least by an all-knowing God). You are asserting that it is theoretically knowable by an all-knowing computer. You are saying (correct me if I'm wrong) that if the computer had access to all my memories, infinite computing power, and all the I/O channels, the computer could predict (and/or replicate) with 100% accuracy, every decision I make. Either way, my subjective experience of being a creative, decision-making entity, capable of choice and responsible for my choices, is obliterated. At best it is a tragic delusion, at worst it is an act of disloyalty to science. If I get sufficently annoyed with you, for instance, to punch you in the nose, what court would convict me? With you as my defence attorney, I'd argue that it wasn't my *fault*, since I *had* to do it, it was all in the programming of the computer in my head. Blame one of the programmers maybe, but not the computer. Responsibility and choice vanish. That is one of the things that makes me very skeptical of the mechanistic theory of human thought. It makes social existence as we have known it for all of recorded history impossible. Of course, this doesn't prove you are wrong, it just suggests quite strongly that you may have overlooked something important. >> AI wants to build machines that can perform tasks or make decisions >>as well as humans. I think though, that human reason and decision >>making is not mechanical, it is a-mathematical, a-scientific and a-rational. >>It is hinged to something else, human values (variable), passion >>(wholly subjective), emotions (volatile) and sympathies (unpredictable). > >Science asserts the exact opposite: That everything can be explained, Hold on a sec mate! Science asserts no such thing. Popular culture asserts that, but few respectable scientists claim anything beyond the capacity to offer some explanations for some things, notably those that can be observed. Popular culture has come to believe that science can tell us how things are, and furthermore, how they must necessarily be. That, sir, is a "religious" enterprise and not a scientific one. A scientist seeks the truth in any way it can be found, and does not presume to know the results before the evidence is in. Science does not assert that "everything can be explained", it merely proceeds to try to explain everything it can see, quantify, and reliably identify. You will note that the history of science is one of revising explanations. The only thing really certain about *every* scientific expalanation is that it is at best incomplete and certain to be modified greatly in time. >that everything can be modelled, and that the way to gain knowledge >is to seek these explanations and models. Nope. Modelling presumes a lot of information, and that presumes an observable phenomenon. Neither modelling nor science presume that "everything" is sufficiently observable to be modelled. You simply dismiss all those things that are not sufficiently observable to be studied scientifically as not existing. That is a very religious assertion, and not at all scientific. >Religion, on the other hand, is completely the opposite. Religion >asserts that there are things that cannot be explained, that some things are >beyond comprehension, and to seek explanation is to lose the faith. Part 1 of that is true. Most theistic religions recognize that there are some things which cannot be explained. Both you and "Science" recognize that too. The search for an explanation is not an anti-religious act. It may be anti-authoritatarian, but most great religious leaders are notable for the fact that they sought, and claimed to have found, higher levels of explanations. Both religion and science seek to know, to explain, and to make experience meaningful. >In other words, your statements imply that you are a member, >unwittingly or no, of a religion whose basic tenate is that there is at least >one thing in the universe (the human mind) which cannot be explained. Just >as other religions have as their basic tenate that the one unexplainable >thing is "God." Hmm. Odd what my statements might imply. Actually I came to computer science late in life. My formal training is in theology, history and psychology. Biblical studies was my major. The basic tenet is not at all that God is unexplainable, but rather that God is knowable. >> created and not instrinsically creative. My hunch is that human >>thought is really dependent on dimensions of the universe which science >>(as we currently understand it) is not yet capable of fathoming. > >I would be interested in hearing the reasoning behind your "hunch". >If you have any evidence or experience that may be relevant to the AI >community, perhaps we should hear of it? I'd point you to the "100th Monkey syndrome" for starters. It's a well known, well documented case of apparent cause-effect realtionships occuring between living creatures with no explanation. Indeed, it is "scientifically" impossible :-). There are lots of other human experiences worth looking into, as a class of phenomena. One that comes vividly to mind is the case of the R.A.F pilot whose navigation instruments broke down over the North Sea in 1956. He attempted to get the attention of coastal radar operators by flying a "distress pattern". As he hoped, an aircraft arrived to "shepherd" him back to his base. For 20 minutes he flew wingtip to wingtip with the "shepherd" during which time he had opportunity to note that aircraft's markings, and that it was a Mosquito fighter-bomber of WW II vintage. After his safe landing he sought out the pilot of the "shepherd" to thank him, only to discover that the R.A.F had no Mosquitos in service any longer and that the plane which had shepherded him to base had disappeared over the North Sea 12 years previously. Science has trouble with things like that becuase it is a non-repeatable phenomenon. You can't really deal with it scientifically very well at all. You can't "prove" much of anything about it, except that a highly trained observer (the pilot) was scared to death and was praying energetically that the radar operators (who were not on duty, it was Christmas Eve) would see his distress pattern (which they did not) and send up an aircraft (which they did not) to rescue him from nearly certain death. You can't really prove that a Mosquito showed up, except that not only the pilot, but the controller at the airfield both reported having seen it. There are more scientific "impossibilities" in the story. The Mosquito could not have aided the instrument landing in the deep fog because the airstrip where the pilot landed did not have electronic navigation aids. The controller heard low-flying aircraft, thought there might be an emergency and switched the runway lights on. The story gets more bizarre actually. But it is one of many which suggest that our ordinary conscious way of apprehending reality might, just might, not be all there is to apprehend. Finally, look into Karl Jung's theory of the collective unconscious, which was developed after a vast amount of very scientific research concerning phenomenon similar to those mentioned above. All of these present challenges to a mechanistic world view. You can dismiss them as nonsense, as Galileo was dismissed, if you want. You can develop lots of hypotheses to "explain" such phenomenon, and that is all worth-while. But an hypothesis is not a fact, remember that. >> How can you apply math or science to such things as Faith, >Spiritual sensibility, relgious experience, love or hatred? > >Oh boy. Want me to give you a long anthropological and/or sociological >treatise upon the relation between faith in shamans, and survival >in primitive societies? Love as an outgrowth of the nurturing instinct which >assures survival of offspring? Hatred as an outgrowth of the warding >instinct which protects the resources of the family group thus aiding >survival? Hmmm. I will return to my first argument. In saying that "love" has survival value you have really said nothing at all. You have certainly said nothing about why I love A and do not love B. You will probably not be very persuaded by that, but I'd say that's because you refuse to deal subjectively with your own experience which is the only device by which you can *know* what it *means* to be a human being, as opposed to being a mechanistic epiphenomenon. >There has been much scientific research into the areas that you >mention. In many of them, we do have some knowledge of the evolution of >various social more's from ancient times to the present. Almost invariably, >the root cause is some behavior which, at that time, had some positive >survival value. Again you are over-stating your case. Establishing some association with evolutionary survival value does not prove a "causal relationship". It merely shows some relationship. Your logic is badly flawed on that one. Basically, you can't get to those conclusions from the evidence at hand without adding a great deal of faith and extrapolation. Your hunch may be right, but it is only a hunch, not a proof -- nor even a theory really. >An assertation that such behaviors cannot be scientifically explained >is thus contradicted by the fact that we have sciences called History, >Anthropology, Sociology, and Psychology. <long low whistle> Well, now we're on more familiar terrain. Sure historians offer "explanations", some of which almost sound believable. Few of them are the stuff of which a mechanistic world-view can be constructed. E.g., why did England not send Wellington's army, flush with vicotry over Napoleon, to the US to finish off the dispute of 1812-14 and teach those insolent colonists a lesson once and for all? Shall I tell you or shall you tell me? The "science" of history offers several hundred different, and sometimes contradictory explanations of the event. (or lack of an event in this case). So tell me, what is *the* scientific explanation of the failure of the UK to prosecute the war of 1812 with more vigour? Why, after burning the White House (which was then red brick) did the British/Canadian armies withdraw? Would you accept the Duke of Wellington's own personal account of the matter? Would you accept it as "scientific" that the war with Napolean had turned the man into a pacifist who never wanted to see another gun fired as long as he lived? Anyway, there are a thousand explanations, and the ones people tend to look at are the ones they like. The typical explanation in US history books is that American armies had given the British a hard enough time they lost their taste for war. The typical Canadian history book recounts "betrayal" of Canadian interests by the Colonial Office. Official British histories tend to look at "exhaustion" after the expensive conflict with Napolean. So which one is "scientific"? >> Science can do very well with the natural world, but I suspect >there is a part of the human being which is strongly connected to a >super-natural reality which science has yet to get a grip on. > >Here is where I become certain that you are advocating a certain >set of religious beliefs (commonly called "New Age", I believe). Sorry, simply good old fashioned Christian :-) >As I said before, the whole underpinning of science is that the world can >be explained. And as I've said before that reveals a very superficial understanding of what science really is, and has been through the ages. Science deals with that which can be explained and has nothing at all to say about anything else. >That approach has yielded every scientific advance in the world today, >from the wheat upon your table (a scientifically-bread hybrid, high yield), >to the car you drive to work. There is no reason to expect that paridigm >of reality to stop yielding results any time soon. Again you are firing off non-sequiturs. Of course Science has done much to help mankind gain power over the natural world. Such things (especially in recent centuries) have been pretty noticable and spectacular. But you betray a serious ignorance of history when you fail to realize the immense contribution to civilized social organization made by the Judeo-Christian tradition. You probably do realize that without a somewhat stable, somewhat decent and somewhat just society, you do not have an enviroment which is conducive to the detached intellectual inquiry upon which modern science (and civilization generally) depends. >On the other hand, the >opposite paridigm (that there are things that cannot be explained, and to >search is to lose your faith), has yielded no results at all besides >innumerable books full of moral advice that everybody seems to ignore. Everyone does not ignore it sir. Quite the contrary, everyone obeys it. If they did not, life would be nasty, brutish and short and we'd still be living in mud huts. Thous shalt not kill. Thou shalt not steal . . . Far from teaching that the cosmos was unexplainable, Christianity and Judaism have taught that the cosmos is reasonable, purposive, and committed to the well-being of mankind. They have taught that to be human is to be important and to be meaningful. Without the philosophical foundation of those ideas and the social forms which grew out of them, such as justice and reason, I'd argue that science would never have been invented. I do not think we are dealing with "opposite paradigms" at all. I think (excuse my arrogance) that my paradigm includes and accommodates yours quite handily, while yours can't fit mine in at all. My paradigm suggests that this is because yours is not big enough. What makes me most pessimistic about the evangelists of AI is the pathetic lack of understanding of what intelligence and the cultures which produce intelligence really consist of. At this point I realize that when I use the word "intelligence" and when you use it, we may not be talking about exactly the same thing. Indeed, we may not have experienced it as the same thing. To hazard a guess, I'd suggest that intelligence is something like the manifestation of wisdom based on a humble appreciation of the lessons of history and culture with recognition of the limits of human life in the natural world. The basic limit there is the grave. We are mortal. We gaze beyond our known limits to the future on the other side of the grave, the unknowable. My guess would be that you see intelligence more as something manifested wherever an individual's behaviour can be show to augment his marginal utility over against the rest of the natural world. You would dismiss afterlife as something about which nothing can be known and therefore nothing can be thought. So I end up having to ask you a question. Would you consider Jesus to be a manifestation of intelligent life? Would you consider me to be a manifestation of intelligent life? (not that I'm trying to put myself in the same league as Jesus). If, as you seem to suggest, the only intelligent way to look at things is wholly mechanistic, then indeed I agree, that kind of intelligence can be mechanized, at least potentially. It is the other kind of intelligence which the machines cannot be taught. Anyway, thanks for your thoughful response. I've enjoyed your thought-provoking challenge. Best regards, Doug Thompson ------------------------------------------------------------------------ Fido 1:221/162 -- 1:221/0 280 Phillip St., UUCP: !watmath!isishq!doug Unit B-3-11 Waterloo, Ontario Bitnet: fido@water Canada N2L 3X1 Internet: doug@isishq.math.waterloo.edu (519) 746-5022 ------------------------------------------------------------------------
ken@cs.rochester.edu (Ken Yap) (06/09/88)
Here is more food for thought: If/when an intelligent machine is created, how do you know it won't have "human" traits, including being lazy, having emotions, getting bored, etc? By intelligent, I don't mean something that is expert in one domain like cleaning your house, but a machine that you can plonk out in the street and it will survive (subject to qualification). I suspect that a machine complex enough to be intelligent will also have other traits. The unspoken wish of AI is to have machines do our dirty work and not complain. I think if you make a machine that smart, it might object to doing dirty work or throw tantrums. Predictability: in principle yes a machine is predictable but if it is so complex that you have to take into account the configuration of the universe at every moment, then it is, for all practical purposes, unpredictable. Already we have complex systems that are capable of surprises, just read comp.risks. Do you think an intelligent machine would be more tractable? Or shall we move to less mechanical ways of describing behaviour, like what novelists do with people? Ken
bzs@BU-CS.BU.EDU (Barry Shein) (06/10/88)
If we do accomplish the basics to build androids I wonder if companies will employ "psycho-artists" to design interesting personalities for them? Will it become a home hobby? (No, that's not right, he still seems peevish not cute, wipe out the empath eproms and we'll try again tomorrow...) Imagine what it could do for the entertainment industry! I agree with Ken, it might be very hard if not impossible to separate human foibles from human strengths in the end, they might be one and the same. Highly complex AI will almost certainly require a lot of self-organization (learning.) Of course, androids will be slaves we won't mind horse-whipping, of course...? -Barry Shein, Boston University
gcf@actnyc.UUCP (Gordon Fitch) (06/10/88)
In article <10431@sol.ARPA> ken@cs.rochester.edu (Ken Yap) writes: } Here is more food for thought: } } If/when an intelligent machine is created, how do you know it won't } have "human" traits, including being lazy, having emotions, getting } bored, etc? .... } I suspect that a machine complex enough to be intelligent will also } have other traits. The unspoken wish of AI is to have machines do our } dirty work and not complain. I think if you make a machine that smart, } it might object to doing dirty work or throw tantrums. The only reason to build a machine which acts just (or pretty much) like a human being is to get rid of human beings, or at least the need for them. So, though there is motivation on the part of some people to do this, while there is an ample supply of human beings most researchers will find themselves trying to build machines which are intelligent in different ways than human beings. We have already seen some of this in the computer field, where "computation" was developed beyond the ability of even the most efficient idiots-savants. Possibly, machines will be built to identify, as well as design, other machines with abilities and goals completely beyond human comprehension. Further machines will then be needed to explain the former to human beings, or determine whether they have become psychotic -- an occupational hazard in these areas.
joe@inria.UUCP (Joe arceneaux ) (06/10/88)
Eric Green writes:
1) The human mind consists of mechanism (program) and data (memories),
I think this maybe imposing the computer science frame of reference a
bit strongly. There are a lot of philosophers who would disagree, and
this is important because, in my opinion, many (most) of the big questions in
this domain are META-physical.
this into an AI model, and ask a machine to decide which is "best" or
which is "right", the democratic party or the republican party. Try to
replicate the human decision-making process at the ballot-box.
Under the above model, your choice is dictated by your past experiences
(memories) and by various mechanisms operating upon those memories (e.g.
mechanism: pleasure/security. Memories: Political discussions, which policies
were best, who was the better speaker, smarter, etc.).
That is, your choice consists of neurons firing in a particular pattern. And,
if somehow, microseconds before the vote, we had the sum total of your
knowledge and experience available to us, we could predict exactly how you
would vote. You could have voted in no other way besides that dictated to you
by your knowledge and experience and the underlying mechanisms.
I suspect rather strongly that people of Werner Heisenberg's ilk would
disagree with you here. Remember that Einstein *never did* show that
"God doesn't play dice with the universe."
Science asserts the exact opposite: That everything can be explained, that
everything can be modelled, and that the way to gain knowledge is to seek
these explanations and models.
Whether it is indeed true that everything can be explained is another
question and once again in the philosopher's ballpark.
Science can do very well with the natural world, but I suspect there is
a part of the human being which is strongly connected to a super-natural
reality which science has yet to get a grip on.
Here is where I become certain that you are advocating a certain set of
religious beliefs (commonly called "New Age", I believe). As I said before,
the whole underpinning of science is that the world can be explained. That
approach has yielded every scientific advance in the world today, from the
wheat upon your table (a scientifically-bread hybrid, high yield), to the car
you drive to work. There is no reason to expect that paridigm of reality to
stop yielding results any time soon. On the other hand, the opposite paridigm
(that there are things that cannot be explained, and to search is to
Well, the statment "yet to get a grip on" is important. There are
certainly myriad phenomena which defy rational explanation. But just
as it is wrong to say "Oh, look! Mystical/Magical/Religious
phenomena!" it is also wrong to assume that eventual explanations of
said phenomena will be satisfying, or even fit into the scheme now
used to [scientifically (:-)] model reality.
Joe
"Use the Source, Luke."
tom@PHOENIX.PRINCETON.EDU (Thomas C Hajdu) (06/12/88)
working towards designing technology with more or less human atributes addresses the future (of humans) issue sideways. without the possibility of fusing humans (biologically?) with technology, humans will become as obsolete as crystal radios.
elg@killer.UUCP (Eric Green) (06/13/88)
In message <53.22AB6402@isishq.UUCP>, doug@isishq.UUCP (Doug Thompson) says: >>From: elg@killer.UUCP (Eric Green) >>So I have little doubt that the human mind IS succeptable to >>scientific analysis. As to whether it can be modelled (part 3), however, > >And therein I think you have expressed the problem clearly. You have >little doubt. It was Descartes who said that the only thing really >certain to him was his own capacity to doubt. It is critical to knowing >anything. I'd refer you to the history of scientific revolutions, in >which we find *doubt* is the primary engine of creativity and new discovery. Actually, the driving force behind scientific revolutions has been EMPIRICAL EVIDENCE. That is, a man may doubt that the sun orbits the Earth, and he may believ that the Earth orbits the sun. But, if he does not collect evidence, it remains just that: a belief, no more valid than the belief that the sun goes around the Earth. There has been evidence that the mind may be analyzable by scientific methods. There has been no evidence of the pseudo-religious assertation that there a quality of the mind (call it a "soul") which renders it unanalyzable by scientific methods. Therefore..... [tons of pseudo-religious babble about Religious Detirminism, Mechanistic detirminism, etc., deleted, much to relief of net I'm sure, who don't feel like listening to Doug be insulted by the thought that he's just a bunch of neurons firing and thus "does not have free will":] With that out of the way: I do not believe that we will ever have a machine intelligence which closely resembles human intelligence. A large part of our human intelligence is derived from experiences processed by our quite effective senses, and from factors such as mobility, growth, the ability to grasp objects, etc. A machine intelligence, fed on information typed into consoles or digitized via cameras or whatever, most probably would have a much different "world view", and thus might be almost incomprehensible to "normal" intelligences. With that out of the way, though, AI MAY eventually pay off. Currently we use computers to mechanize tasks which are too repetitive and too linear for our relatively slow (~50khz) brains to process at any reasonable rate of speed (but, while we may be slow, we sure are parallelized). AI could result in computers capable of handling similiar tasks that require more "intelligence". For example, while there are currently database services that will look up references based upon keywords, usually you get either too many references to look through, or you restrict the search too far and miss relevant articles (some of which may be quite important). One can envison describing what kind of criteria you're interested in, and having the computer pick exactly the same papers as you'd pick if you were doing it the old-fashioned way :-). As for "naughty computers": I can't voluntarily stop breathing even if I wanted to. After I topple over unconcious, I start breathing again. It's possible to envision a computer that needs to do database searches the same way (hmm, a bit extreme of an example, but you get the point).
ken@cs.rochester.edu (Ken Yap) (06/13/88)
A couple of posters have pointed out that machine intelligence may be quite different from human intelligence. I fully agree with the possibility but would like to add that we didn't rigously define the word "intelligence" (and may not be able to reach agreement on this). So while you may be thinking of smart news reading machines, I may be thinking of androids. Certainly machine intelligence may not need to emulate human thought, just as jet planes work on a different principle from bird flight. But let me reverse the analogy and say both the plane and the bird have to obey the laws of aerodynamics. Perhaps we will discover that the kind intelligence that develops is is partly governed by (goeswith) the environment. So a Canopean being may have different, maybe even incomprehensible ways of thinking from us. Already we have problems with cross-culture understanding. So maybe a electronic intelligence that inhabits "imaginary" worlds of information may be comprehensible to only similar beings. But I think machines that have to work closely with humans may (have to) take on human characteristics and weaknesses. The idea of the indefinite perfectability of humans is an older idea than AI. How much better a place the world would be, we think, if everybody was a perfect reasoning being. There would be no human foibles to cause wars, etc. We think that with AI perhaps we can create machines that would embody pure intelligence and no faults. I think that for ourselves and our creations, both assets and liabilities will spring from the same source. Sounds like I should be posting to rec.sci-fi instead. :-) Ken
charlesv@auvax.UUCP (Charles van Duren) (06/13/88)
In article <4347@killer.UUCP>, elg@killer.UUCP (Eric Green) writes: > in article <48.22A3B84F@isishq.UUCP>, doug@isishq.UUCP (Doug Thompson) says: > > Science asserts the exact opposite: That everything can be explained, that > everything can be modelled, and that the way to gain knowledge is to seek > these explanations and models. > > Religion, on the other hand, is completely the opposite. Religion asserts that > there are things that cannot be explained, that some things are beyond > comprehension, and to seek explanation is to lose the faith. > > In other words, your statements imply that you are a member, unwittingly or > no, of a religion whose basic tenate is that there is at least one thing in ^^^^^^ > the universe (the human mind) which cannot be explained. Just as other > religions have as their basic tenate that the one unexplainable thing is > "God." Where to begin. Your posting leaves me with the desire to write tomes on the incredible myopia of (much) of the AI community, and of practically all of the advocates of the scientific method. The scientific method is a religion. It is circular to argue that the scientific method has more merit or truth or whatever because its effectiveness can be demonstrated (by the scientific method, of course). Don't mistake me. I am not denying the value of the scientific method. All I am saying is that, as a method, it is only a few centuries old, so let us keep it in perspective. > There has been much scientific research into the areas that you mention. In > many of them, we do have some knowledge of the evolution of various social > more's from ancient times to the present. Almost invariably, the root cause is ^^^^^^ > some behavior which, at that time, had some positive survival value. > How narrow and impoverished your view of life must be! > An assertation that such behaviors cannot be scientifically explained is thus ^^^^^^^^^^^ > contradicted by the fact that we have sciences called History, Anthropology, > Sociology, and Psychology. > Attempts, with indifferent success. Let's not get carried away. Sane persons will integrate the knowledge gained from these endeavours in their world-view, but a pinch of salt is recommended. > Here is where I become certain that you are advocating a certain set of > religious beliefs (commonly called "New Age", I believe). As I said before, > the whole underpinning of science is that the world can be explained. That > approach has yielded every scientific advance in the world today, from the > wheat upon your table (a scientifically-bread hybrid, high yield), to the car > you drive to work. These are _physical_ phenomena. Also, the wheel was invented sometime before the scientific method, so don't generalize about the scientific method. Isn't it convenient to be able to label another person's views and dispose of them so easily. > (that there are things that cannot be explained, and to search is to lose your > faith), has yielded no results at all besides innumerable books full of moral > advice that everybody seems to ignore. Allow me to say it. What bullshit. You generalize from insufficient, no, non-existent data or experience (a scientific no-no). For many people of faith, the search never stops. For many others, the search (and I don't mean one limited by the rules of scientific inquiry) _is_ the faith. > The stumbling blocks to AI are much more mundane: the > large bulk and limited computing power of today's computers, providing enough > input devices to provide a suitable store of memories upon which to operate, > etc. > Yea, yea, I've heard it all before. Technology will provide the answers for the problems of technology. Let me reiterate. I do not expect that my views will have any effect on your thinking. I would love to have the time to develop them more fully for you, but that will have to wait. The scientific method, applied to what is _human_ (yes, I do believe that we are different from machines), leads to a critically impoverished view of our lives and futures. If you believe that we are mere machines, or that our behaviour can be explained in mechanistic terms, I will fight you on the beaches, etc., because this is an unacceptable invasion of my very human life. Charles van Duren (human)
dc@gcm (Dave Caswell) (06/15/88)
In article <672@auvax.UUCP> charlesv@auvax.UUCP (Charles van Duren) writes:
.Where to begin. Your posting leaves me with the desire to write tomes on
.the incredible myopia of (much) of the AI community, and of practically
.all of the advocates of the scientific method. The scientific method is
.a religion. It is circular to argue that the scientific method has more
.merit or truth or whatever because its effectiveness can be demonstrated
.(by the scientific method, of course). Don't mistake me. I am not denying
.the value of the scientific method. All I am saying is that, as a method,
.it is only a few centuries old, so let us keep it in perspective.
.Attempts, with indifferent success. Let's not get carried away. Sane persons
.will integrate the knowledge gained from these endeavours in their world-view,
.but a pinch of salt is recommended.
How does one treat knowledge with a pinch of salt? Do you sometimes ignore
it, sometimes follow it, or what? Moreover reality does not care whether
you have integrated its facts into your worldview or not; they will still
go on being facts. It isn't something I would even care to argue; the
world makes it clear. And then you go on to say that your whole point
is that the scientific method is relatively new; what's the point, finally
a partially defensible position to fall back on?
.> Here is where I become certain that you are advocating a certain set of
.> religious beliefs (commonly called "New Age", I believe). As I said before,
^^^^^^^
.> the whole underpinning of science is that the world can be explained. That
.> approach has yielded every scientific advance in the world today, from the
.> wheat upon your table (a scientifically-bread hybrid, high yield), to the car
.> you drive to work.
.These are _physical_ phenomena. Also, the wheel was invented sometime before
.the scientific method, so don't generalize about the scientific method.
Do you mean to imply that no one did science or followed the scientific
method until one specific date in the past? Did rocks also often float
upwards until the theory of gravity was formalized; what an absurd thought.
There is a difference between an invention and a discovery. Wheels and
science aren't inventions in the same sense.
If you weren't sure that this was merely some religious dogma(though
I'm not sure that New Age describes it) you can be sure now.
The search for what? What are the tools for the search?
.futures. If you believe that we are mere machines, or that our behaviour
.can be explained in mechanistic terms, I will fight you on the beaches, etc.,
.because this is an unacceptable invasion of my very human life.
But don't you realize that beliving that people are articles of faith is
having a much lower opinion of them. People are thinking, reasoning beings
and to degrade reason is to degrade reason and knowledge is to degrade
humanity itself.
--
Dave Caswell
Greenwich Capital Markets uunet!philabs!gcm!dc
If it could mean something, I wouldn't have posted about it! -- Brian Case
dc@gcm (Dave Caswell) (06/15/88)
In article <672@auvax.UUCP> charlesv@auvax.UUCP (Charles van Duren) writes: >Allow me to say it. What bullshit. You generalize from insufficient, no, >non-existent data or experience (a scientific no-no). For many people of >faith, the search never stops. For many others, the search (and I don't >mean one limited by the rules of scientific inquiry) _is_ the faith. I think that the whole thrust of his posting lies in the few lines I quote above. Maybe you could explain what the search is for, what you search with (the tools for your inquiry). How do you carry the search out? How is the search helped by not being so limited by the rules of scientific inquiry? What is the result of the search, is it knowledge or something else? What does "the search is the faith" mean. I know that you have point to make but sometimes the narrowness of the crt screen makes discussion difficult. Elaborating on your point will make things much clearer. I think that you are trying to present the alternative to science but you don't quite finish the explanation. If you have even more time explain, what is a "person of faith". People who aren't "of faith", what are they? In others words, what is the opposite? If I wanted to be a person "of faith" what would I have to do or not do? Is it a desirable to be of faith; would the world be a better place if all of us were people of faith. If I have misunderstood and everyone is a person of faith, I apologize not really knowing what being "of faith" means in this context. If I look in my dictionary for faith I see Belief that does not rest on logical proof or material evidence: faith in miracles. And for miracle it says superantural or an act of God. And for supernatural it says a miracle. So you can well belive that after a definition like that, that some more explanation from you would be helpful. Thanks. -- Dave Caswell Greenwich Capital Markets uunet!philabs!gcm!dc If it could mean something, I wouldn't have posted about it! -- Brian Case
michael@stb.UUCP (Michael) (06/16/88)
In article <4347@killer.UUCP> elg@killer.UUCP (Eric Green) writes: [Discussion of predicability of humans, and modeling by computers] >To generalize, there are a couple of basic assumptions needed to make >AI a science instead of a religion: > >1) The human mind consists of mechanism (program) and data (memories), >2) All human action is detirmined by the operation of the mind's mechanisms > upon the mind's memories, >3) The computer can model the above. > >In particular, > [ specifics for points 1 and 2] > >> this into an AI model, and ask a machine to decide which is "best" or >> which is "right", the democratic party or the republican party. Try to >> replicate the human decision-making process at the ballot-box. > >Under the above model, your choice is dictated by your past experiences >(memories) and by various mechanisms operating upon those memories (e.g. >mechanism: pleasure/security. Memories: Political discussions, which policies >were best, who was the better speaker, smarter, etc.). > >That is, your choice consists of neurons firing in a particular pattern. And, >if somehow, microseconds before the vote, we had the sum total of your >knowledge and experience available to us, we could predict exactly how you >would vote. You could have voted in no other way besides that dictated to you >by your knowledge and experience and the underlying mechanisms. Careful of what can of worms you open. Allow me to re-phrase you: If we knew exactly what program you operate on, and the data availible to you the moment before you enter the booth, we know exactly what you will do. If we knew exactly what program you operate on, and the data availible an hour ago, and the inputs you received during that hour, we know exactly what you will do. If we knew exactly how the world was to the decimal, and we know your state at that same moment, then we can calculate all the inputs you will receive and determine exactly what you will do at any time in your life. Congradulations! You have just denied the existance of free will. This is a highly debatable point. Determinists hold your position, that everything can be determined exactly. (Quantum mechanics seems to indicate the opposite, that everything is random and statistical at best. Personally, I'm hoping that a third posibility will be discovered that allows free will and freedom of choice.) >> AI wants to build machines that can perform tasks or make decisions as >> well as humans. I think though, that human reason and decision making is >> not mechanical, it is a-mathematical, a-scientific and a-rational. It is >> hinged to something else, human values (variable), passion (wholly >> subjective), emotions (volatile) and sympathies (unpredictable). > >Science asserts the exact opposite: That everything can be explained, that >everything can be modelled, and that the way to gain knowledge is to seek >these explanations and models. Really? We can't even say what an explanation is (i.e., no specific step-A step-B step-C "this is an explanation" algorithim exists that agree with our intuition). Try modeling a single electron sometime--tell me where it will be. >Religion, on the other hand, is completely the opposite. Religion asserts that >there are things that cannot be explained, that some things are beyond >comprehension, and to seek explanation is to lose the faith. Why can't we go faster than the speed of light? "Well, its just a fundamental limit" But why is it a fundamental limit? "It just is" I hope you don't think I'm nit picking here, but... > >To say that artificial intelligence cannot be attained because of magical or >mystical properties of the human mind is thus of no relevance to scientists. >Similiar statements have been made about other scientific endeavors, and have >always proven false. The stumbling blocks to AI are much more mundane: the >large bulk and limited computing power of today's computers, providing enough >input devices to provide a suitable store of memories upon which to operate, >etc. Actually, I have a bigger stumbling block to propose: Turing machines can do anything a major computer can do. There are a countable number of turing machines, but an uncountable number of problems. Therefore, any computer will be limited in what it can do. A computer cannot do things that it is not programmed to do. Humans, however, have hunches, can deduce the existance of a double helix spiral, or a wave-particle combination, without having had it programmed in to them. The point: Beyond a certain point, all you can program a computer to do is try every possible combination til something looks right. Humans have hunches, guesses, flashes of insight, etc. and check only a few ideas before comming close. > >-- > Eric Lee Green {cuae2,ihnp4}!killer!elg > Snail Mail P.O. Box 92191 Lafayette, LA 70509 >"Is a dream a lie if it don't come true, or is it something worse?" Michael p.s. Followup should probably go to a philosophy of science group, but I don't know of any. : --- : Michael Gersten uunet.uu.net!denwa!stb!michael : ihnp4!hermix!ucla-an!denwa!stb!michael : sdcsvax!crash!gryphon!denwa!stb!michael : What would have happened if we had lost World War 2. Well, the west coast : would be owned by Japan, we would all be driving foreign cars, hmm...
ken@cs.rochester.edu (Ken Yap) (06/16/88)
The scientific method is not merely about finding answers to yes/no questions. Yes, the earth is not flat, but then it is not quite a sphere either. It is an oblate spheroid, with bumps and depressions due to local variations in gravity. The point here is that the real world is much richer than a collection of binary facts can capture. Those who argue against the possibility of AI on the grounds of that it reduces man to machines are on shaky ground. Think about those who fought to retain a heliocentric universe. So maybe man is a collection of cells but what a wonderful collection, as revealed through biology. Why do you feel insecure when there is simply no comparison? Mysticism has nothing to do with it. Certainly there are other modes of perceiving the world other than the objective one of the scientific method, as well there should be. Science doesn't claim to know everything and does not prove there is no God or supernatural phenomenon, or whatever. (By the way, if your idea of God is a superbeing who put together the earth and its inhabitants from a Meccano set at 4 pm, July 20th 7000 AD, or something like that, perhaps you should take a less restrictive view of God.) Perhaps, in the fullness of time, SOME mysteries will receive rational explaination, but I believe there will always be richer secrets to understand about the universe. Arguing that AI is not possible because machines cannot be mystics is the old "machines cannot do X" argument. I agree with those who rail against the trend to deal with the world in mechanistic terms. Is AI possible? is not an interesting question because it is a yes/no question and will be proved one way or another, assuming we don't blow ourselves up first. The more important question to ask is: what do we WANT it to be like? Years ago I read a tongue-in-cheek article in one of those personal computing rags, the type with Star Trek programs in BASIC, about the log of a burglar alarm system. In short, the system identified the intruder as a recidivistic criminal, received authorization from the central security databank to terminate the intruder, zapped the criminal with a 10 MW laser then proceeded to clean the mess. At least I hope it was tongue-in-cheek, because I find this kind of thinking scary. There are things that are simply not appropriate to delegate to machines. These issues, and those surrounding other technologies like gene splicing bear much discussion before the future is upon us. Ken
ken@cs.rochester.edu (Ken Yap) (06/17/88)
|Meccano set at 4 pm, July 20th 7000 AD, or something like that, perhaps
^^
Oops, you know what I meant. That's what I get for writing stuff at
6 pm in the morning, er, I mean 6 am. :-)
Ken
chris@MIMSY.UMD.EDU (Chris Torek) (06/17/88)
Let me begin by saying that I think this <newsgroup+mailing list> (yes, it is both) is not the appropriate place for this discussion. That out of the way...: >In article <672@auvax.UUCP> charlesv@auvax.UUCP (Charles van Duren) writes: >>... The scientific method is a religion. Let me try the following definitions: religion \equiv faith faith \equiv [from Webster's ?th ? dictionary] 2b1: firm belief in something for which there is no proof 2b2: complete confidence 3: something that is believed esp. with strong conviction; esp : a system of religious beliefs Let me use a form of `3' above, i.e., `something that is believed with strong conviction'. Then: the scientific method is not a religion. (In point of fact, the scientific method is just a method---one of many possible methods, but one that a large number of people believe works.) Belief in the efficacy of the scientific method, however, is (by my definitions) a `faith' and therefore a `religion'. Recall that the only thing for which I have absolute and utter proof is that `I think, therefore I am.' Anything else---any evidence of my senses---I must take on faith. (Perhaps I am the butterfly, dreaming.) >>... Don't mistake me. I am not denying the value of the scientific >>method. All I am saying is that, as a method, it is only a few >>centuries old, so let us keep it in perspective. Make that `as a systematically defined and applied method' and I will agree. (Not, of course, that those who apply it are any less human, or somehow never err. The scientific method is limited not only by what it is, but also by those who apply it.) >>... Also, the wheel was invented sometime before the scientific >>method, so don't generalize about the scientific method. Dave Caswell answers: >Do you mean to imply that no one did science or followed the scientific >method until one specific date in the past? Did rocks also often float >upwards until the theory of gravity was formalized; what an absurd thought. >There is a difference between an invention and a discovery. Wheels and >science aren't inventions in the same sense. This is tangential to what I see as van Duren's point, which is that `real' things (which, unless they are only your own existence, require some measure of `faith') can be invented, discovered, or described by other methods. (I personally believe that the scientific method is the most efficient method that we can ever use. With an appropriate definition of efficiency, one could compare various methods to prove whether it is the most efficient we have yet found, but without a method to characterise all possible methods, my belief must remain a faith.) >>If you believe that we are mere machines, or that our behaviour can >>be explained in mechanistic terms, I will fight you on the beaches, etc., >>because this is an unacceptable invasion of my very human life. >But don't you realize that beliving that people are articles of faith is >having a much lower opinion of them. People are thinking, reasoning beings >and to degrade reason is to degrade reason and knowledge is to degrade >humanity itself. I happen to believe that we *are* machines, though hardly `mere'. But, at least as yet, there is little that science can say about this. We know only that many things that were once claimed to be impossible for machines are in fact possible. Extrapolation to the extent that anything we can do, a machine could do, is just that---extrapolation; and extrapolation without testing does not make good science. Incidentally, you can always look at it this way: You might as well believe in free will. If free will exists, wonderful! You were right all along. And if free will is just wishful thinking, if all our actions are predetermined, why, then, it was not your fault you were wrong: You had to be wrong; it was predestined. :-) -- In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7163) Domain: chris@mimsy.umd.edu Path: uunet!mimsy!chris
cbs@geacrd.UUCP (Chris Syed) (06/17/88)
In article <10526@sol.ARPA>, ken@cs.rochester.edu (Ken Yap) writes: > I think that for ourselves and our creations, both assets and > liabilities will spring from the same source. > Ken Me too. Hegel (who unfortunately majored in obscurity), remarked that the same spirit (or mind) worships in religion and reasons in philosophy (which to him meant all science). I take this to mean that you can't chop up the mind/brain's faculties and come up with isolated modes of thinking that would work on their own. Reading Ken's posting, I got to wondering - what if we tried to make a machine think like, say, a porpoise or a gorilla? I see a connection between that and trying to create one that thinks pure logic. We can certainly make one that _behaves as we see porpoises behaving_, or expert systems that drive railway trains like engineers do.... but will they think like engineers do at the ball game or the symphony? I think that eventually, we can build a "complete porpoise machine", but that still won't solve the basic problem. We don't think pure logic because our perceptions of 'observable reality' are the inputs, and they're integral to the judgements we make about what we see. Flip this around, and you come up with the idea that there ain't no pure logic _that we can know about_ - just human logic, or Canopean or gorilla logic. To 'really' deal with 'objective truth', we or our machines would somehow have to get 'outside' the universe and look in... etc. etc. This is what led Hegel, for example, to his conclusion that 'phenomenology' (the art of reasoning in the world of appearances), was the way to go. {uunet!mnetor,yunexus,utgpu} !geac!geacrd!cbs (Chris Syed)
evan@saturn.ucsc.edu (Evan Schaffer) (06/29/88)
In article <10425@stb.UUCP> michael@stb.UUCP (Michael) writes: >Actually, I have a bigger stumbling block to propose: Turing machines can do >anything a major computer can do. There are a countable number of turing >machines, but an uncountable number of problems. Therefore, any computer >will be limited in what it can do. Actually, there are an infinite number of turning machines. Turing machines may have an infinite number of states. In fact, one can argue that, given the size of a neuron, and the size of a human head, there are a limited number of neurons that will fit in a human's head, so a turing machine is capable of more complex behavior then a human. Michael Wolf ARPA: wolf@ssyx.ucsc.edu UUCP: ...ucbvax!ucscc!ssyx!wolf
bzs@BU-CS.BU.EDU (Barry Shein) (06/30/88)
Although I don't subscribe to vitalism I'm also not sure that there isn't some middle ground here (other than with those who insist on appealing to purely supernatural events, there's really no where to go from that starting point.) For example, physics has certainly been dabbling in this century with very non-deterministic and, importantly, anti-intuitive concepts such as quantum mechanics. This leads one to ask if the mind is not in many ways a probability machine, learning etc simply varying the probability (sometimes to near zero/one) of certain events happening. It might very well be that just as we needed group theory and other approaches before we could even ask the right questions in physics we might very well need a better way/model to talk about the functions of the mind. For example, viewing thinking as an n-dimensional probability space and learning as something deforming that space seems at least intuitively appealing (therefore it's probably wrong :-) It also may turn out that trying to separate anything physical from the conscious is an error and there's very little useful thought that can go on without experiental sensation, there is a continuum between the external and internal (note how disruptive experiences like isolation tanks can be to thinking.) This can be simulated, the notion of an hallucinating program seems apt. Another thought that has occurred to me (and others) is that few AI researchers have the patience (or the grant money!) to put the kind of time into an AI program that your average parent puts into your average child (not to mention the number of bits the rest of the world feeds a child.) -Barry Shein, Boston University
sullivan@vsi.UUCP (Michael T Sullivan) (07/01/88)
In article <3965@saturn.ucsc.edu>, evan@saturn.ucsc.edu (Evan Schaffer) writes: > Actually, there are an infinite number of turning machines. Turing machines > may have an infinite number of states. In fact, one can argue that, > given the size of a neuron, and the size of a human head, there are a limited > number of neurons that will fit in a human's head, so a turing machine is > capable of more complex behavior then a human. Just where does one put a Turing machine with an infinite number of states? One could argue that with an infinite number of neurons, thus implying an infinite head, that a human brain can be infinitely powerful. But of course we are back to the problem of where does one put an infinite head? -- Michael Sullivan {uunet|attmail}!vsi!sullivan V-Systems, Inc. Santa Ana, CA sullivan@vsi.com ons, workstations, workstations, workstations, workstations, workstations, work
bay@arvak.cs.cornell.edu (Paul Bay) (07/01/88)
In article <740@vsi.UUCP> sullivan@vsi.UUCP (Michael T Sullivan) writes: >In article <3965@saturn.ucsc.edu>, evan@saturn.ucsc.edu (Evan Schaffer) writes: >> Actually, there are an infinite number of turning machines. Turing machines >> may have an infinite number of states. In fact, one can argue that, >> given the size of a neuron, and the size of a human head, there are a limited >> number of neurons that will fit in a human's head, so a turing machine is >> capable of more complex behavior then a human. > >Just where does one put a Turing machine with an infinite number of states? >One could argue that with an infinite number of neurons, thus implying an >infinite head, that a human brain can be infinitely powerful. But of course >we are back to the problem of where does one put an infinite head? Turing machines have a *finite* number of states. See Hopcroft and Ullman, "Intro to Automata Theory, Languages, and Computation." In theory, the human brain, since it has only a finite number of neurons, is no more powerful than a finite automaton. In fact, even if you augment the human brain with all the libraries and all the paper in the world, it's still no more powerful than a finite automaton, since these extra resources are finite as well. Reply via email; this discussion doesn't belong in this group.
dm@DIAMOND.BBN.COM.UUCP (07/01/88)
>Actually, there are an infinite number of turning machines. Turing machines >may have an infinite number of states. In fact, one can argue that, >given the size of a neuron, and the size of a human head, there are a limited >number of neurons that will fit in a human's head, so a turing machine is >capable of more complex behavior then a human. >Michael Wolf Actually, as the person to whom you were replying pointed out, there are only a countable infinity of Turing machines (that is, Turing machines can be put into a one-to-one correspondence with the integers). I'll have to be convinced that there is an uncountable number of problems (seems plausible at first blush, but then, countable vs. uncountable infinities DON'T seem plausible at first blush, so I don't like rely on mere plausibility where levels of infinity are concerned). As regards your second point --- that our finite heads imply there are things we can't conceive --- Rudy Rucker explores this idea in great detail in his _Mind tools: the five levels of mathematical reality_. His meditation is inspired by Gregory Chaitin's algorithmic information theory (``You can't get a 20 pound theorem out of a 10 pound axiomatic system''), which is also described in the June, 1988 _Scientific American_ (in the article ``Randomness in arithmetic'').
ken@cs.rochester.edu (Ken Yap) (07/02/88)
|Another thought that has occurred to me (and others) is that few AI |researchers have the patience (or the grant money!) to put the kind of |time into an AI program that your average parent puts into your |average child (not to mention the number of bits the rest of the world |feeds a child.) Hmm. Maybe we should make computers cuter and cuddlier? :-) Ken
bzs@BU-CS.BU.EDU (Barry Shein) (07/02/88)
>|Another thought that has occurred to me (and others) is that few AI >|researchers have the patience (or the grant money!) to put the kind of >|time into an AI program that your average parent puts into your >|average child (not to mention the number of bits the rest of the world >|feeds a child.) > >Hmm. Maybe we should make computers cuter and cuddlier? :-) > > Ken In a way we have, that's why the academic position of "technical typist" is rapidly becoming a thing of the past. The world used to use computers for answers, now we all do data entry...it's almost fun given the right WYSIWYG etc. -Barry Shein, Boston University
ins_atge@jhunix.HCF.JHU.EDU (Thomas G Edwards) (07/03/88)
In article <712@inria.UUCP> joe@inria.UUCP (Joe arceneaux ) writes: > > That is, your choice consists of neurons firing in a particular pattern. And, > if somehow, microseconds before the vote, we had the sum total of your > knowledge and experience available to us, we could predict exactly how you > would vote. You could have voted in no other way besides that dictated to you > by your knowledge and experience and the underlying mechanisms. > >I suspect rather strongly that people of Werner Heisenberg's ilk would >disagree with you here. Remember that Einstein *never did* show that >"God doesn't play dice with the universe." I doubt quantum uncertainty has much effect on unpredictability of the human mind. Chaos, however, does. Each neuron is a non-linear unit, and we have 10^10 to 10^11 of them, which makes the situation so Chaotic that long term predictions (more than a half a second) would be unfeasable. But, the fact remains your hand touched the lever of your candidate. Your hand was positioned by nerves extending back to your brain. Your brain outputed nerve impulses to your hand. Those impulses cam from neurons excited by other neurons' impulses, ad almost infinitum. > Science asserts the exact opposite: That everything can be explained, that > everything can be modelled, and that the way to gain knowledge is to seek > these explanations and models. > >Whether it is indeed true that everything can be explained is another >question and once again in the philosopher's ballpark. It always seemed to me that science examined evidence, and based on that evidence cam up with ways of organizing information and making predictions. Chaos, interestingly enough, says that although you can make micro-level predictions, and it predicts that macro-level predictions of large non-linear systems is very difficult, but that all non-linear systems have certain universal characteristics. -Thomas G. Edwards ins_atge@jhuvms
ins_atge@jhunix.HCF.JHU.EDU (Thomas G Edwards) (07/03/88)
Again, getting back on the "Chaos Express," I think we are going to have to change our ideas of "random" and "probability" for the future. Chaos has shown us that seemingly random results can be derived from very simple non-linear equations (see the mandelbrot set for a fine example). The tenet of Chaos is that small changes in initial parameters can have incredibly huge effect on non-linear systems, even seeming random though _they are deterministic_. Furthermore, it states that due to this, micro level deterministic events creat undpredicatble large level realities. But a mandelbrot set is still the same everytime your run it...so one has determinism, with the addendum that although systems may be predicted in terms of chaotic attractors (general recurring patterns), they may not be predicted in exactness. -Thomas Edwards ins_atge@jhuvms
doug@isishq.UUCP (Doug Thompson) (07/04/88)
elg@killer.UUCP (Eric Green) writes: > >In message <53.22AB6402@isishq.UUCP>, doug@isishq.UUCP (Doug >Thompson) says: >>>From: elg@killer.UUCP (Eric Green) >>>So I have little doubt that the human mind IS succeptable to >>>scientific analysis. As to whether it can be modelled (part >3), however, >> >>And therein I think you have expressed the problem clearly. >You have >>little doubt. It was Descartes who said that the only thing >really >>certain to him was his own capacity to doubt. It is critical >to knowing >>anything. I'd refer you to the history of scientific revolutions, >in >>which we find *doubt* is the primary engine of creativity and >new discovery. >Actually, the driving force behind scientific revolutions has >been >EMPIRICAL EVIDENCE. That is, a man may doubt that the sun orbits >the Quite. But a man only hears what he wants to hear. If you aren't looking for evidence, you don't find it. If someone hadn't doubted that the earth was flat, the evidence of its roundness would never have surfaced. It *was* there all along, and even the Classical Greeks knew the Earth was round. I simply wanted to point out that the mind-set with which you undertake an inquiry into the evidence often influences the evidence you'll find. If you have little doubt, you will find little evidence because you won't be looking for it. ------------------------------------------------------------------ Fido 1:221/162 -- 1:221/0 280 Phillip St., UUCP: !watmath!isishq!doug Unit B-3-11 Waterloo, Ontario Bitnet: fido@water Canada N2L 3X1 Internet: doug@isishq.math.waterloo.edu (519) 746-5022 ------------------------------------------------------------------
fbaube@NOTE.NSF.GOV (Fred Baube) (07/06/88)
Thomas Edwards: > .. The tenet of Chaos is that small changes in initial > parameters can have incredibly huge effect on non-linear systems, > even seeming random though _they are deterministic_. > Furthermore, it states that due to this, micro level deterministic > events creat undpredicatble large level realities .. Can the precipitating/determining events themselves be caused deterministically if they're operating at a quantum level ? Couldn't observable macro events be determined (chaotically) by probabilistic events ?
ins_atge@jhunix.HCF.JHU.EDU (Thomas G Edwards) (07/08/88)
In article <8807051413.aa01556@note.note.nsf.gov> fbaube@NOTE.NSF.GOV (Fred Baube) writes: > >Thomas Edwards: >> Talks a little about Chaos > >Can the precipitating/determining events themselves be caused >deterministically if they're operating at a quantum level ? >Couldn't observable macro events be determined (chaotically) >by probabilistic events ? As far as quantum determinicy goes, I am not sure on the relevant points. I believe the going hypothesis is that quantum events are deterministic, yet unpredictable (perhaps chaos at a very small/fast level?) But yessiree, very small quantum changes are just the kind of nasty things that can drive chaos, and now that I've officially hurt my head thinking about this, we can now safely say that quantum events may play a nasty part in the stochastic selection of whether a neuron fires or not, because the chemical transfers across the synapse are probably at least slightly chaotic. But remember, chaos does exists on computers...just try plotting the Mandelbrot Set, Henon Attractor, or the Xnext=rx(1-x) equation. -Thomas Edwards