Newsgroups: comp.ai.games
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!Germany.EU.net!EU.net!sun4nl!hermes.bouw.tno.nl!usenet
From: sst@bouw.tno.nl (Tako Schotanus)
Subject: Re: The turing test
Message-ID: <1995Mar22.115011.16629@hermes.bouw.tno.nl>
Sender: usenet@hermes.bouw.tno.nl (USEnet Postmaster id)
Nntp-Posting-Host: ruudnix
Organization: TNO Bouw
X-Newsreader: WinVN 0.92.6+
References: <3k9p54$j05@newssvr.cacd.rockwell.com> <3kd50e$sp9@newsbf02.news.aol.com> <3kkdf3$dab@newssvr.cacd.rockwell.com> <3kko4r$6pi@vixen.cso.uiuc.ed <3kl0r8$id5@newssvr.cacd.rockwell.com>
Date: Wed, 22 Mar 1995 11:50:11 GMT
Lines: 208

In article <3kl0r8$id5@newssvr.cacd.rockwell.com>, csmccue@cacd.rockwell.com (Craig S. Mc Cue) says:
>
>Just reposting an E-Mail I got from Mickwest
>
>---------------------------------------
>
>
>csmccue@cacd.rockwell.com (Craig S. Mc Cue) wrote:

[Stuff deleted]

>As for Consciousness, the above thought experiment could be conducted between a 
>conscious person and an unconscious person. As for sentience, I think this is a 
>philosophical term, and not capable of the straightforward analysis given 
>above. As a rule of thumb, I would say that animals are conscious, but not 
>sentient, wheras humans are both. Consciousness implies the ability to relate 

Would you include primates?

>to the outside world, and includes instincts that help the animal to survive in 
>the outside world. Conscious animals can communicate, socialize, build houses 
>(like nests and beaver dams), use tools (like the otter using a stone to crack 
>a nut open), fight for survival, plan out a search for food, etc. Sentient 
>beings are able to do the above, but also seem to be able to distinguish 
>between the outside world, and an inner, subject world of their own making. 
>Sentient beings are able to symbolize, imagine, create and intuit relationships 
>between objects. They can be driven by desires stronger than instinct, and can 
>ignore instinct altogether for higher principles. It can believe in something 
>greater than itself -- something which it can strive for.

Which could also be things we just like to BELIEVE are the "great" things that
distinguish as from animals. Take creation for example: they always take the
"arts" as an example that "we" humans are far above animals because we have
art, a totaly useless way to spend your time were it not for the fact that we
sometimes like to look at/listen to the results of that waste of time :)
What if it turns out to be, after years of research into the human brain,
that the "creative gift" is nothing more than an "urge" to (re-)combine
experiences in novel (maybe random/chaotic) ways and apply them to something
you're good at (writing, painting, singing, etc.). Maybe the main difference
between a poet and one of those random lyric generators is just the human
experience, the knowledge how our society works and therefore being able to
predict wether a certain frase or sound will sound pleasant to somebody else.
A computer not having that experience has no way to "cull" his random "creations"
and therefore no way to improve himself.

I do think your "defintion" of sentience is biased towards humans, what we
should come up with is a more general definition that would include aliens
from planet X as well. (Or primates and computers from the next century as well)

NB: An example of the belief in our supereority: religious trance is said by some
people to be the highest state of mind our kind can reach. Some have claimed that
experiments show that this state is comparable to people using certain kinds of
drugs (If I remember correctly LSD was one of them, but I'm not sure) and that all
that the drugs did was a partional shutdown of the cortex. The effect this could
have is: you stop "thinking" about what you're doing/feeling. What's left is just
the feeling. Which is how some people describe "trance": you just stop thinking
and start feeling. But if it's the cosmos you're experiencing or just the rustling
of an empty mind..... who knows...



>>Supposing a computer can simulate human conversation and reasoning to such an 
>>extent that I cannot distinguish it from a human conversing and reasoning.
>>
>>I don't know what is going on inside the computer, I don't know what is going 
>>on inside the human, my external perception of the two is the same.
>>
>You may not be able to (legally) dissect a human to see what is going on 
>inside. But you can examine the computer to see if what is going on inside its 
>processor is the same thing as what is going on inside your own mind.
>

True, but not true. Some complex neural networks can do simple things humans
can. Researchers thought "hey, we can take the net apart after it has
learned what it had to learn and we can figure out how it works and maybe
learn how the human brain does the same thing in the process!" The problem
was that the resulting neural net was just as hard to understand as the human
one!

>>What is the difference. We have two active things, one made of flesh and 
>>blood, the other of rare metals. Both act in the same way. We have no way of 
>                                          ^^^^^^^^^^^^^^^^^
>as of yet, there are no computers that can act like human beings.
>>following their internal reasoning and we expect it to be radically different, 
>>but we observe their external responses are very similar.
>
>As stated above, yes we can dissect their innerds, as it were, and observe how 
>they think (ever hear of a debugger?)

You do NOT want to run a debugger on a neural network program! Do you know how
many parallel "processes" you'd have to be able to follow. I think even the
greatest human minds would have problems trying to do that.

NB: By processes I don't mean computer-type processes, just that all the 
inter-relating signal being sent between all the different neurodes can be
seen as small seperate processes that are going on simultaniously.

>
>>Clearly what goes on inside is all the difference. Is it a mechanical 
>>difference? Are humans concious because of their fuzzy complex brains? If you 
>>could open up a human head and replace the neurons and synapses in there with 
>>some silicon equivalent, cunningly interface by some advance in 
>>bio-engineering .. if you could do that a little, would the person whose brain 
>>it was become less intelligent if you replaced a little of their brain with 
>>identically functioning silicon? What if you replaced the whole brain?
>
>How much of consciousness has to do with instincts? The instinct to survive, to 
>breath, to eat and expunge, to procreate, to socialize, to fill wanted, to love 
>and be loved? Computers have no such instincts -- even if you were able to 
>incorporate 'pseudoinstincts' into a program, the program would only function 
>as programmed, not because it needed to eat. Can a computer ever love? Would 
>acting as if it loved another mean that it actually did? Human beings act as if 
>they love someone else, without truly loving them, all the time.

Why would a computer need love? We're not trying to create pseudo-humans are we?
Would an intelligent-sentient alien from planet X EVER understand what love was?

>

>>>If an F/A-18 1553 simulator appears to be flying to the subsystems it is 
>>>talking to, then it IS flying (something wrong here).
>>
>>Bad analogy.
>>
>>"Flying" is somthing well defined, meaning moving rapidly through the air. 
>>"Inteligence" on the other hand, is only defined in terms of the way it 
>>manifests itself. The mechanics of flying are well understood and it is 
>>perfectly obvious that the simulator is not actually flying. The mechanics of 
>>intelligence are hardly understood at all, and we do not know if it is the 
>>same from person to person. If a computer seems to be intelligent, by what 
>>authority can we say that it is not intelligent?
>
>I agree -- for the case of intelligence. But intelligence does not imply 
>consciousness or sentience. We can test a computer's intelligence by the 
>criteria I gave above. If it meets that criteria, then we can call it 
>intelligence (but intelligence of a computer-kind, not of a human-kind). I 
>think the analogy is appropriate, because it shows that something appearing to 
>have a certain quality does not mean it DOES have that quality. I'll use a more 
>ephemeral example: A Pharisee APPEARS to be righteous. From his actions we see 
>many righteous deeds, and many good works. But IS he righteous? He may have 
>motives of political power, or may be hiding an evil intent. He may practice 
>unrighteous acts in private, where no one can see him. Put another way, because 
>Jimmy Swaggart cries on TV, does that mean he has actually repented, or is only 
>putting on an appearance of repentance?

Lots of people seem to believe he does. So what happens when we can't agree
ourselves? To many people he has actually repented and if nothing ever
happens to convince them otherwise....


>It appears to be the case that the sun rises in the east and sets in the west. 
>From our relativistic vantage point, we can assume that the sun travels around 
>the earth. Or we can formulate a copernican model and explain the reality 
>another way: the earth moves around the sun. Neither one is reality, however. 
>They are merely models useful for us. In actuality, the sun, the earth, and 
>everything else in the solar system travels around a Centroid, which is the 
>actual center of the solar system (and which does NOT correspond to the center 
>of the sun).
>
Ah ah ah! A model is just that: a model, it has no "reality", it just reflects
reality as we know it. Your "centroid" is just another way of looking at it
(hey, it doesn't even exists :) and it's just as true or just as false as the
(pre-)copernican models. The only real difference between those models, the one
that caused so many problems, as a theological one: is earth the center of 
everything or isn't it? Finaly, it all depends on your point of view: you
could *still* use the earth-centered models, there's nothing wrong with them
(a bit over-simplified nowadays, but that could be changed :).
So appearances are enough. For all I care it the "truth" that the sun rises in
the east and sets in the west. So it IS that way. My model is enough for what I
want, it's not enough however if I want to talk with an astonomer. But his model
of reality isn't more "truthful" than mine, it just how things APPEAR to be to
him and for him that's enough, for him that's the way it IS.

>>Can you give me an example of something appearing to be what is not? If so, 
>>then how can you tell that it is not what it appears to be? By something other 
>>than appearance?
>
>I just did... 
No you didn't: you just replaced one APPEARANCE with another :)

>Bottom line: you assume that your experience of a phenomena 
>reveals all that it can about that phenomena. This ignores the possibility that 
>other, more accurate means can be detected, which can shed more light on a 
>subject, and which can allow us to model reality more closely to how it 
>actually is.

No no, for a model it's just necesary to be *correct* it doesn't have to be *complete*!

If a model "works" for you it's enough: if I work with clay to make pottery or
whatever I don't need the physical nor the chemical models of clay to be able to
make something. So somebody might come up with a replacement for clay made from
plastics for all I care and if it behaves the way clay does I might call it clay
(especially if they replaced it without me knowing it) on they other hand it might
be obvious that it is not clay but I can still use it for whatever I want, so it
might not be clay *technically* but if I have no physics/chemistry background
I might not care.

> Truth is only an illusion, albeit a very good 
>>one.

Everything's an illiusion, but as long as we don't notice the diffence we can say
that everything IS as it APPEARS to be :)

_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
_/ Tako Schotanus                TNO Building and Construction Research _/
_/ Phone : +31 15 842393 Fax : +31 15 122182  E-mail : sst@bouw.tno.nl  _/
_/ My employer is required,by Dutch law,to disagree with whatever I say _/
_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
