About
Content
Store
Forum

Rebirth of Reason
War
People
Archives
Objectivism

Post to this threadMark all messages in this thread as readMark all messages in this thread as unreadBack one pagePage 0Page 1Page 2Page 3Page 4Page 5Page 6Page 7Page 8Page 9Forward one pageLast Page


Post 40

Saturday, August 13, 2005 - 6:02pmSanction this postReply
Bookmark
Link
Edit
Jordan,
...But computer can count, so under your test, they would be rational.
The test doesn't define the ability to count as rationality, all of rationality and nothing but, it merely claims that rationality is a requirement for completing the test.

That neither you or a stone can fly doesn't prove that you should be a stone, failing the test does not prove lack of rationality.

A computer is an inanimate object, created by rationality to perform a specific task - counting - it will not 'know' what counting is, or what to do with it - it is it's programmed reaction. Birds will need to transcend their programmed reaction to obtain this ability, they will need some form of rationality, if they fail the test they may still be rational, just not proven by this test.

I do though hold that man is an animal like most others, we have more complex reactions but i don't think it fair to say that we react in wildly different ways to the same stimuli just as i don't find it fair to say that animals always react with similar responses to same stimuli.

An animal i have had a chance to observe, would be my cat at home... it would often sleep in my bed, and want to leave sometime during the night, it would need me to open the door. It started by going to the door meowing, if that did't work it would walk across my piano, on the keys, if still no result it would go to the door moving its paws across the metal slats of the venetian blinds - no claws just the paw - a terrible noise, and it would normally get my attention, one night when even this failed, it placed its front legs on the edge of my bed and ever so gently nibbled my nose - no pain involved except that of waking up to the foul odor of cats breath. It had learned that it needed my attention to get the door open, it had learned that noise was a good way, it found that when noise failed it could resort to other means. Not that this particular cat was exceptional, but i think it showed rational thought on some level... abstraction.

At the same time, it was claimed that birds always build the same type of nest, while humans clearly don't... don't we? Don't we use the materials at hand to build very similar houses - more complex than birds nests, and with greater variety sure - we are more complex - but amazingly unique i would not call our dwellings. And for birds... when at the local hospital last, on the balcony outside my window was a bird that had build its nest of steel 'sticks' - those used to reinforce brick walls, and bind inner walls to outer walls.... it was the material it could find at the site - the nest was made of nothing but that.

If we look for animal cognition while accepting one result only, be it that they are rational or not, it is futile to search any further, as we already know that our result will be what we have decided it to be.

That a gorilla can't be the king of france is silly - the gorilla will need to be seen in gorilla world, and humans in human world - that a gorilla wont become the king of france - let us pretend france a monarchy for a moment - is just as obvious as the fact that you won't become king or alpha male of a mountain with different groups of silverbacks.

If we look at the great diversity of life on our little globe, we see that many living entities, however different we may appear, have an awful lot in common, two eyes, two ears, a nose, feet and hands, hearts, lungs, livers, brains ... that humans should be the only animal capable of forming abstract concepts i would find highly unlikely, that we don't have the mental capacity to prove what goes on in the brains of other animals, may in fact say more about us than it does them.

Post 41

Sunday, August 14, 2005 - 1:39amSanction this postReply
Bookmark
Link
Edit
Soren, your post was directed at Jordan, but yet it takes issue with me (and what I've written) -- so I'll respond.

--------------
The test doesn't define the ability to count as rationality, all of rationality and nothing but, it merely claims that rationality is a requirement for completing the test.
--------------

Right. It takes rationality for an unprogrammed being to count beyond that which is perceptual. Evidence of counting up to the limit of that which is perceptual is best explained by memory and crude association (perceptual powers of awareness). And evidence of counting beyond that which is perceptual -- is lacking.


--------------
i don't think it fair to say that we react in wildly different ways to the same stimuli just as i don't find it fair to say that animals always react with similar responses to same stimuli.
--------------

How about the thousands of "wildly different" languages we've developed -- some involving several senses of verbs, and some few; some involving several senses of pronouns, and some few; some involving a single word that stands for a whole sentence in other languages; some involving a change in meaning from mere inflection of intensity (without a different atomistic pronunciation); and some involving "clicking" sounds, mixed in with verbalizations?

Any given species of animals, on the other hand, uses the same methods of communication -- there is no observable change across space or time. From the beginning of written history, dogs have always barked, and cats have always meowed -- and there is no indication that this will not continue with the exception-less regularity that it always has.


--------------
It had learned that it needed my attention to get the door open, it had learned that noise was a good way, it found that when noise failed it could resort to other means. Not that this particular cat was exceptional, but i think it showed rational thought on some level... abstraction.
--------------

The best explanation of the behavior of your cat is that it was using signals (concrete, perceptual) and not signs (abstract, conceptual).


--------------
At the same time, it was claimed that birds always build the same type of nest, while humans clearly don't... don't we? Don't we use the materials at hand to build very similar houses ... ?
--------------

Have you missed the work by F. L. Wright? How about the tents of American Indians, or the igloos of the eskimos? Do I need to continue? Do you see that there is nothing at all like this in nature? There is no species of animal (other than humans) that demonstrates this potentiality for variability in its behavior -- this ability for a single species to find a million unforeseen ways to get the same things done?


--------------
when at the local hospital last, on the balcony outside my window was a bird that had build its nest of steel 'sticks' - those used to reinforce brick walls, and bind inner walls to outer walls.... it was the material it could find at the site - the nest was made of nothing but that.
--------------

It was the material it could (perceptually) find at the site -- that is the best explanation. If something looks (to a bird) like a twig, then it will be treated as a twig -- much the same phenomenon as dogs barking at themselves in the mirror.


--------------
that humans should be the only animal capable of forming abstract concepts i would find highly unlikely, that we don't have the mental capacity to prove what goes on in the brains of other animals, may in fact say more about us than it does them.
--------------

No comment.

Ed



Post 42

Sunday, August 14, 2005 - 6:39amSanction this postReply
Bookmark
Link
Edit
Soren,

I took Ed's test as meaning that if something can count, then it's rational. This would make computers rational, a conclusion I thought Ed wouldn't want to draw. I recognize that counting is a sufficiency test rather than a necessity test, so I understand that other beings who fail to count might still be rational. But that's just it. Sufficiency tests really won't help us get to the heart of the problem. Too many beings will fail the counting test, leaving us to wonder whether those great many beings are rational. What we could really use is a necessity test.

Ed,
How about the thousands of "wildly different" languages we've developed
All communicative humans use the same general grammatical rules. In that sense, we're no different from animals. But really, if all humans did use the same language (like in some pre tower of babylon community), would you really deny them rationality?
There is no species of animal (other than humans) that demonstrates this potentiality for variability in its behavior -- this ability for a single species to find a million unforeseen ways to get the same things done?
Many animals are exceptional problem solvers. They learn, often systematically, changing their behavior again and again in order to achieve their goal. Ever see a squirrel get at at a birdfeeder?

Jordan


Post 43

Sunday, August 14, 2005 - 8:59amSanction this postReply
Bookmark
Link
Edit
Jordan,

-----------------
I took Ed's test as meaning that if something can count, then it's rational. This would make computers rational, a conclusion I thought Ed wouldn't want to draw.
-----------------

Change "something" to "some unprogrammed thing." Things that are programmed to count -- don't count here.


---------------
What we could really use is a necessity test.
---------------

Thompson 3-pronged Necessity Test for Rationality

1) identifiable use of First Principles
2) capability of dealing with Unfamiliar Particulars -- and in Unprecedented Ways
3) identifiable aims toward Final Ends


---------------
But really, if all humans did use the same language (like in some pre tower of babylon community), would you really deny them rationality?
---------------

If all humans used the same language, then Pinker would actually be on to something -- enter innate mental content.


---------------
Many animals are exceptional problem solvers. They learn, often systematically, changing their behavior again and again in order to achieve their goal. Ever see a squirrel get at at a birdfeeder?
---------------

Yeah ... so! I've seen rats run mazes, too. Big deal. Rats can run mazes because rats get hungry, look for the food, and REMEMBER where it wasn't. Just because a rat can "learn" to remember a complex maze, doesn't make it a conceptualizer.

Once, there was a chimp that made a tool -- an extendo-stick -- from 2 sticks in order to get at bananas outside a cage. The clincher is that the chimp had to PERCEIVE the 2 sticks as one (on the ground) when they came in line with each other. Chimps don't ever do the unperceivable.

Ed



Post 44

Monday, August 15, 2005 - 7:58amSanction this postReply
Bookmark
Link
Edit
Ed,
Change "something" to "some unprogrammed thing." Things that are programmed to count -- don't count here.

Why not? If counting is sufficient for rationality, then shouldn't it be sufficient in each case? If not, then counting is neither necessary or sufficient for rationality.
Thompson 3-pronged Necessity Test for Rationality

1) identifiable use of First Principles
2) capability of dealing with Unfamiliar Particulars -- and in Unprecedented Ways
3) identifiable aims toward Final Ends
We've been over this. Your 3 prongs aren't empirically meaningful until you come up with some concrete way to measure them.
If all humans used the same language, then Pinker would actually be on to something -- enter innate mental content.
The big deal with innate mental content was popularized by Chomsky, who found that the general rules for grammar are universal. I think his conclusion was right; his reasoning (i.e., about innate mental content), wrong.

Last, you wrote:
There is no species of animal (other than humans) that demonstrates this potentiality for variability in its behavior -- this ability for a single species to find a million unforeseen ways to get the same things done?
That some animals solve problems systematically - i.e., demonstrate variability in their behavior - refutes this view. I wasn't trying to argue that problem solving gives rise to conceptual capacity, just that problem solving gives rise to variability in behavior.

Gotta run.
Jordan


Post 45

Tuesday, August 16, 2005 - 5:25amSanction this postReply
Bookmark
Link
Edit
Jordan,

--------------
If counting is sufficient for rationality, then shouldn't it be sufficient in each case? If not, then counting is neither necessary or sufficient for rationality.
--------------

Jordan, integrate this with my 3 prongs. For those sentient, counting is a sufficient test for sapience. For the circuit-boards and microchips, the 3-pronged approach becomes necessary.

2 different types of agents -- 2 different approaches for sufficiency.


--------------
The big deal with innate mental content was popularized by Chomsky, who found that the general rules for grammar are universal. I think his conclusion was right; his reasoning (i.e., about innate mental content), wrong.
--------------

I'd love to see your "correct" reasoning on this. The only thing innate is method and means (which has to do with the human way of gaining knowledge -- which, itself, has to do with our unique brains, etc). Initial content is provided by perception, and perception only.

Later -- after some perceptions -- a mind may juxtapose past perception into "imagination" (e.g. centaurs: from man and horse; mermaids: from woman and fish). Also, the human mind may begin to think of things (e.g. justice) that are not at all instances of perceptions.

Ed

Post 46

Tuesday, August 16, 2005 - 7:35amSanction this postReply
Bookmark
Link
Edit
Hi Ed,

2 different types of agents -- 2 different approaches for sufficiency.
I understand your point; I'm just not sure it's a valid move. I thought you applied the counting test to both humans and animals, two different types of agents. Why seperate robots and humans but not animals and humans? See, I guess if counting is indicative of rationality, then if something demonstrates counting, we should accept that it has rationality. If counting is not indicative, then it shouldn't be used. If counting is indicative of rationality in only a select few, then you have to explain (via other tests) the reason for such limited application.

--------------
The big deal with innate mental content was popularized by Chomsky, who found that the general rules for grammar are universal. I think his conclusion was right; his reasoning (i.e., about innate mental content), wrong.
--------------

I'd love to see your "correct" reasoning on this.
Hmmm. Maybe we have a misunderstanding. I reject "innate mental content" and agree with you that only methods and means are innate. Chomsky amassed a ton of evidence to show that humans everywhere share general rules of grammar. I think Chomsky is right that we share universal grammar, but as I understood it, he originally attributed this universality to innate mental content, a view a disagree with. I think it's more supportable to explain universal grammar by appeal to our innate methods and means. I hope that clears things up.

Jordan



Post 47

Tuesday, August 16, 2005 - 8:26amSanction this postReply
Bookmark
Link
Edit
Jordan, the litmus test for whether counting is sufficient is whether sentience is affirmed.

Ed


Post 48

Tuesday, August 16, 2005 - 9:20amSanction this postReply
Bookmark
Link
Edit
Ed,

So if a being is sentient, then if that being can count, then it's rational? Again, it would help if you justified why counting is sufficient to indicate rationality only for sentient beings. Further, now you'll need to come up with another (empirical) test to determine sentience.

Jordan

(Edited by Jordan on 8/16, 9:20am)


Post 49

Tuesday, August 16, 2005 - 12:59pmSanction this postReply
Bookmark
Link
Edit
Jordan,

Poke 'em with a pin. If they shriek and jump in obvious fear -- they're sentient.

Ed


Post 50

Tuesday, August 16, 2005 - 2:04pmSanction this postReply
Bookmark
Link
Edit
Ed,

Funny. So now I have to not only be able to shriek and jump and do it when poked with a pin in order to qualify for the counting-as-entailing-rationality test? Friggin weird, man.

I could quibble that I don't think shrieking and jumping after pin-pricks aren't necessary for pain, and that pain isn't necessary for sentience, but I don't want to digress. Instead, I'll just say you gave me another sufficiency test, which, like the other sufficiency tests, leaves too great a number of unknowns in that everone who fails the test might still have the thing we're looking for.

Jordan


Post 51

Friday, January 6, 2006 - 1:37pmSanction this postReply
Bookmark
Link
Edit
Jordan,

A necessity test would involve learning-as-you-go. It is precisely something that can't be captured by a pre-written program. Think Godel, here. Sure, I already hear you shriek: "You've merely defined computers out of the possibility of being rational, by claiming that nothing pre-written could capture rationality!"

Yup.

Ed


Post 52

Friday, January 6, 2006 - 1:41pmSanction this postReply
Bookmark
Link
Edit
Yup indeed - and that has always been the problem with AI...

Post 53

Friday, January 6, 2006 - 3:46pmSanction this postReply
Bookmark
Link
Edit
Ed:
A necessity test would involve learning-as-you-go. It is precisely something that can't be captured by a pre-written program.
Why not?

Post 54

Saturday, January 7, 2006 - 11:07amSanction this postReply
Bookmark
Link
Edit
Cal, I'll quote Mortimer Adler on the matter. Note: Adler and I acknowledge the Turing (conversation) Test as evidence of rationality ...

My reason for thinking that AI machines will never pass this test is that the many turns in an extended human conversation are unpredictable and what is unpredictable cannot be programmed. Even when computers are programmed to act in certain random ways, the degree and character of that randomness is programmed. But an extended human conversation has a randomness and unpredictability that is unprogrammable. Hence no AI machine will ever be built that can be programmed to pass the conversation test.
Ed

source:
http://radicalacademy.com/adlercomputers.htm


Post 55

Saturday, January 7, 2006 - 11:42amSanction this postReply
Bookmark
Link
Edit
In other words, Ed - that's you am conversing with, not 'brainiac', huh... ;-)

Post 56

Saturday, January 7, 2006 - 12:51pmSanction this postReply
Bookmark
Link
Edit
Reverend, I've been programmed to respond thusly, and cannot do otherwise:

[gangsta' rap-style]
Yo, yo, yo, -- my brother-from-another-mother!
What' up, Gee? Jus' tell me how you be?
I got 'dis programmer yo, and he keep tellin' me what ta' be sayin'
Why can't a brother speak his mind? Yo, yo, yo -- 'dis is "human race-ism!"

[chorus]

Ed


Post 57

Saturday, January 7, 2006 - 1:45pmSanction this postReply
Bookmark
Link
Edit
Cal,

[from Premoral Choice to Live thread]

But why should only a choice by a being with conceptual power of awareness be a "real" choice?
Because only conceptual beings have their own ideas. And "free" choice involves the stubborn or willful adherence to one's own ideas.

The choices that animals have to make are just as real.
Does a beaver "choose" to build a dam? Does a spider "choose" to spin a web? Are there such things as web-weaving spiders -- who "choose" not to? No.

whether it is that of the animal with its more limited power of awareness (though calling it "only [perceptual]" is an unwarranted simplification, at least for "higher" animals)
So Cal, you are one of the Brutes Abstract Pack. Do you have evidence pertaining to this conclusion of yours? If so, post it. Let's examine it together. You will find that I can be very reasonable, when the situation calls for it.

It's only while we can't predict what that choice will be that we call it "free will".
This does seem intuitively true, I'll give you that. When "others" make unpredictable choices, then we do ascribe a free will to them -- but there is another way to get to evidence of free will, namely, introspection. The realization that, while you are focusing on it, you have control of your mind. You, like me, and everyone else -- can choose to blank out at will. This is often how prisoner's or others deal with extreme punishment or adverse circumstances -- by willing a more pleasant, "internal" environment, be it a blank out, imagery, or willful hope.

Ed


Post 58

Saturday, January 7, 2006 - 1:49pmSanction this postReply
Bookmark
Link
Edit
Michael K.,

[An animal] can choose whether to use its other survival faculties in a given situation - and even how and to what extent to use them.
Can a spider choose how to spin its web? Or does a given species of spider have the exact same web, from spider to spider, generation to generation -- woven perfectly on the very first try?

Ed


Post 59

Saturday, January 7, 2006 - 2:44pmSanction this postReply
Bookmark
Link
Edit
My reason for thinking that AI machines will never pass this test is that the many turns in an extended human conversation are unpredictable and what is unpredictable cannot be programmed. Even when computers are programmed to act in certain random ways, the degree and character of that randomness is programmed. But an extended human conversation has a randomness and unpredictability that is unprogrammable. Hence no AI machine will ever be built that can be programmed to pass the conversation test.

This is a mere assertion that computers can't pass the Turing test, without any real argument ("...that is unprogammable". Why? Blankout, to paraphrase Rand). It seems that Adler had a rather naive view of what AI entails. I get the impression that he thought that it would be a program with a large amount of canned responses, with some (pseudo)random function thrown in. But such a program has little or nothing to do with real AI (although people have been fooled by such programs). He had probably no notion of self-learning programs/networks and genetic algorithms. No doubt AI has still a very long way to go, probably we should think in terms of decades, but I still have to see a substantial argument that it can't be done.

Post to this threadBack one pagePage 0Page 1Page 2Page 3Page 4Page 5Page 6Page 7Page 8Page 9Forward one pageLast Page


User ID Password or create a free account.