From newshub.ccs.yorku.ca!torn!utcsri!rpi!zaphod.mps.ohio-state.edu!cs.utexas.edu!uunet!ogicse!pdxgate!dehn!erich Wed Aug 12 16:52:02 EDT 1992
Article 6539 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!torn!utcsri!rpi!zaphod.mps.ohio-state.edu!cs.utexas.edu!uunet!ogicse!pdxgate!dehn!erich
>From: erich@dehn.mth.pdx.edu (Erich Boleyn)
Newsgroups: comp.ai.philosophy
Subject: Super-intelligence? (was -> Re: Defining intelligence)
Message-ID: <erich.712198847@dehn>
Date: 27 Jul 92 01:00:47 GMT
Article-I.D.: dehn.erich.712198847
References: <BILL.92Jul14224037@ca3.nsma.arizona.edu> <1992Jul15.233344.6478@u.washington.edu> <BILL.92Jul1201712@ca3.nsma.arizona.edu> <1992Jul18.034521.27041@u.washington.edu>
Sender: root@pdxgate.UUCP
Lines: 130

sanelson@milton.u.washington.edu (S. A. Nelson) writes:

>In article <BILL.92Jul16201712@ca3.nsma.arizona.edu>
>	bill@nsma.arizona.edu (Bill Skaggs) writes:

>>sanelson@milton.u.washington.edu (S. A. Nelson) writes:
>>
>>   I'm with the Turing-Test crowd in that I think the closest thing 
>>   we have to defining "X is intelligent" is "X behaves like me."
>>
>>If this were true, the sentence "X is far more intelligent than any
>>living human" would seem absurd to us. (We would interpret it as
>>"X behaves far more like me than any living human".)  It doesn't seem
>>absurd to me -- it seems quite correct to say that God, as described
>>in the New Testament, is far more intelligent than any living human.
>>(God in the Old Testament I'm not so sure about.)

   This brings an interesting related point to light.  To a certain
extent, most (if not all) of what we have to measure/compare "intelligent"
activity by is either the human organism (and human culture throws more than
a few twists ;-)...  and for the ambitious researchers out there, we can
look at other primates, but little else empirical at this point provides any
valuable measure.  Now also consider that we are dealing with a very small
primary learning period, even compared to existing archived information,
such as books.  Please bear with me if the next few paragraphs don't
seem to make much sense, but they do tie into a few points...  there are
main ones I'll tie together.

   1st thread: That said, consider the following thought-experiment...  assume
for the sake of argument that one came across a human that had lived (in a
fairly active mental state) for 200 years so far...  (let's also assume that
neuronal degeneration which normally happens in older humans has not taken
place, and yet assimilation of information/etc. still proceeds along what
could be considered "normal" lines for a fairly young person...  although
who knows what would *really* happen in such a case).  Anyway, it seems
reasonable to assume that such a being would likely have developed far
enough psychosocially such that behavior patterns would either be very
ingrained or very complex...  likely both if living in a stimulus-rich
environment.  Would the behavior of such a person be very easily interpretable
by "human" standards?  What happens when we deal with one who is 300, 400,
say, 1000 years old?  Would such behavior be interpretable at all in any
but its most simplistic modes?  Might we just be hopelessly primitive by
such standards?

   2nd thread: How do you (yes, you, the reader! ;-) remember parents, older
friends, and such when much younger (especially in the range 6-12, for
example), at least for myself, I had a way of dealing with them that involved
more of a model of empirical experimentation vs. goals desired than any kind
of real sense of how they would respond on many issues (yes, at 10 I was
"going on 40" ;-).  Simply, however much they may have been different than
myself in interests, outlook, or "raw intelligence", they had so much more
experience that I had to resort to guessing games of what they were after,
partially since their experience effected their *motives*.  I was in a
situation which I was very unprepared to deal with.  One could argue that
this is a physical maturity argument, more than an age argument...

   Notice that in none of the above was I speaking of anyone qualitatively
"more intelligent" than a human...  between humans there seem to be
qualitative differences in "intelligence", although most of it seems to
be quantitative differences.  Also, I referred mostly to complexity of
social and goal-directed behavior...  since these seem to guide, if not
define, the direction/interpretability of bahavior to a large extent...
although I am certainly treading on my own particular bias here.

>	Well, religion is mysterious. The God of the New Testament is not
>someone you're just going to run into at K-mart, only smarter. God is
>interesting because we can't actually rationally concieve of someone that
>smart. Arrogant as it sounds, I do think it's absurd (in a philosophical
>sense) to speak of something smarter than humans. i.e: we can't make sense
>of it. I've never heard anyone but a hippy say "That slug could be much
>more intelligent than us. You can tell because it doesn't act anything like
>us." And anyway, isn't God considered smarter than us because He acts the
>way we're *supposed* to?

   (at the risk of being flamed, psychology and Occam's Razor argues that
Sunny is likely correct in his suggestive conclusion about religion, and
particularly Judeo-Christian derived forms, in the above paragraph)

   3rd thread: I can't say I agree that it's *absurd* to speak of something
smarter than humans, although one must be very careful *how* ones speaks
about such a being/thing.  A very interesting sci-fi book (for those inclined
to philosophical meanderings) is Stanislaw Lem's "Solaris".  It speaks of
an ocean (yes, a whole ocean, covering most of the planet) exhibiting
bahavior (at least it seems far too unlikely to be random occurance)
that is so complex and unrepeating as to be beyond confusing into the almost
totally unclassifiable.  Although the author seems to make some assumptions
about how people would themselves behave under those kind of conditions,
it provides an interesting and well-done perspective.  One, IMHO,
particularly insightful point is how scientists seemed to continually try
to find a model under which any particular behavior would be classified.
They tried and tried, but never seemed to succeed, particularly since the
"ocean" never seemed to respond in any simple way to experiments, and
would never exhibit repeatable responses for long to any stimulus.  They
kept backing off the levels of phenomena that they were trying to capture
with a model until all they could do was give names to a bunch of events
whose overall patterns seemed repetitive, although it was known that the
specifics never seemed to repeat.

   Tying the separate examples/points together to some extent, consider
how a "500-year-old" human would respond to experiments to determine if
s/he were intelligent?  Likely s/he would get bored very very quickly.
Intelligence tests on Dolphins seem to exhibit this characteristic.  At
least a "500-year-old" might still have enough interest (or enough
curiousity about us) to go through such a test...  but, IMHO, is would not
be particularly meaningful.  A human very much older, or one with a
qualitatively "much greater intelligence" would be even harder.  How well
would an "intelligence test", or just trying basic communication, work
on someone "super-intelligent"?  Beyond some fuzzy boundary, one can
imagine communication of any meaningful kind breaking down entirely.  At
that level, it would probably a comparison in ontological sophistication of
a human adult speaking to a tiny baby.  Pretty much anything of importance
or interest to such a being would be just beyond our capabilities to
conceive of in the slightest.

   Of course, these anologies are inaccurate in the same sense that
predictions of future technology are.  Each jump to a "next-higher"
qualitative state seems to be a *very* non-trivial generalization of the
last stage.  Thomas Khun is a good reference for the beginnings of the
modern versions of these kind of theories.

   Well, I could go on, but it would be polite to let someone tell me just
how wrong I am at this point ;-).

   Erich

--
             "I haven't lost my mind; I know exactly where it is."
    / --  Erich Stefan Boleyn  -- \       --=> *Mad Genius wanna-be* <=--
   { Honorary Grad. Student (Math) } Internet E-mail: <erich@dehn.mth.pdx.edu>
    \  Portland State University  /      WARNING: INTERESTED AND EXCITABLE


