From newshub.ccs.yorku.ca!torn!utcsri!rutgers!sun-barr!ames!ncar!noao!amethyst!organpipe.uug.arizona.edu!organpipe.uug.arizona.edu!bill Tue Jul 28 09:41:20 EDT 1992
Article 6446 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!torn!utcsri!rutgers!sun-barr!ames!ncar!noao!amethyst!organpipe.uug.arizona.edu!organpipe.uug.arizona.edu!bill
>From: bill@nsma.arizona.edu (Bill Skaggs)
Newsgroups: comp.ai.philosophy
Subject: Re: Defining other intelligence out of existence
Message-ID: <BILL.92Jul14102805@ca3.nsma.arizona.edu>
Date: 14 Jul 92 17:28:05 GMT
References: <1992Jul8.092458.3088@otago.ac.nz> <1992Jul10.202045.23753@dcs.qmw.ac.uk>
	<BILL.92Jul13114604@cortex.nsma.arizona.edu>
	<1992Jul14.031930.3423@mp.cs.niu.edu>
Sender: news@organpipe.uug.arizona.edu
Organization: ARL Division of Neural Systems, Memory and Aging, University of
	Arizona
Lines: 66
In-Reply-To: rickert@mp.cs.niu.edu's message of 14 Jul 92 03: 19:30 GMT

rickert@mp.cs.niu.edu (Neil Rickert) writes:

   Bill Skaggs writes:
   >
   >As a prototype, think of a dog standing in front of a fence, with a
   >bone on the other side; twenty feet away the fence has a hole.  The
   >problem is to reach that scrumptious bone; the initial state is the
   >dog's current position, transitions are movements, and there is a
   >single goal state, namely the bone's position.

     Aren't you doing just what the subject line of this article is trying
   to avoid - namely "defining other intelligence out of existence?" 

My intention was to do exactly the opposite.  When the dog solves this
problem, it is showing intelligence.  I don't understand why you say
this is a homocentric view of intelligence -- it seems completely
species-independent to me.  The only thing required to speak of the
intelligence of an entity is that it have goals; the class of
goal-possessing entities contains a lot more than just humans, doesn't
it? 


     I would much rather look at intelligence as something which has evolved.
   Thus I would measure intelligence in terms of a creature's ability to
   adapt to a broad variety of circumstances, since surely this adaptability
   is one of the forces in the evolution of intelligence.

Well, we already have a perfectly good word for the ability to adapt,
namely "adaptability".  It seems wasteful to use the word
"intelligence" to mean the same thing.

But you've weakened my argument in any case.  I was proposing "the
ability to solve problems" as a *descriptive* definition of
intelligence, meaning that I thought this was pretty close to the way
most people usually use the word.  You use the word differently; if
most people do, then I am wrong.

   [ . . . ]

   With situational intelligence, you are getting closer to the idea
   of adaptability.  But it is important to remember that the appropriate
   situations for an alien creature may be totally different from the
   situations we would use for ourselves.  Thus, if you use problem solving,
   you may run into the difficult that there is no measuring rod which can
   measure both human intelligence and alien intelligence.  Perhaps a
   situation might arise in which we would solve a particular problem to deal
   with the situation, but our alien might come upon an alternative approach
   of finessing his way around the situation so that that same problem does
   not need to be solved.  Who is to say one approach is really more intelligent
   than the other?

I agree completely.  Situational intelligence is a set of strategies
for dealing with particular kinds of problems.  In some cases we may
be able to say that one strategy is more "powerful" or "sophisticated"
than another, but it is unreasonable to try to order all strategies on
a strict linear scale.  In other words, I don't believe it makes any
sense to look for an "species-independent IQ test".

Intelligence, abstractly viewed, always involves search and pruning.
The power of a searching mechanism can be quantified, but no
interesting problem can be solved by pure search (because of
combinatorial explosion), so any interesting intelligence requires
pruning, and there is no general way of quantifying the power of a
pruning mechanism.

	-- Bill


