From newshub.ccs.yorku.ca!torn!cs.utexas.edu!asuvax!ncar!zaphod.mps.ohio-state.edu!caen!destroyer!uunet!sequent!muncher.sequent.com!bfish Wed Aug 12 16:52:15 EDT 1992
Article 6555 of comp.ai.philosophy:
Newsgroups: comp.ai.philosophy
Path: newshub.ccs.yorku.ca!torn!cs.utexas.edu!asuvax!ncar!zaphod.mps.ohio-state.edu!caen!destroyer!uunet!sequent!muncher.sequent.com!bfish
>From: bfish@sequent.com (Brett Fishburne)
Subject: Re: Defining other intelligence out of existence
Message-ID: <1992Aug3.235017.21685@sequent.com>
Followup-To: comp.ai.philosophy
Sender: bfish@sequent.com
Nntp-Posting-Host: sequent.sequent.com
Organization: Sequent Computer Systems Inc.
References: <980@engcon.marshall.ltv.com>
Date: Mon, 3 Aug 92 23:50:17 GMT
Lines: 55

In article <980@engcon.marshall.ltv.com> ropella@engcon.UUCP (GEROPELLA) writes:
>In
>Article 7010 of comp.ai.philosophy:
>Subject: Re: Defining other intelligence out of existence
>Date: 16 Jul 92 15:45:48 GMT,
>
>From: bfish@sequent.com (Brett Fishburne) writes:
>
>me:>Until then, I suggest you stick with artificial (emulatable, simulatable,
>me:>and analyzable) intelligence.  (Which would be communication patterned
>me:>after ours.)
>
>>I don't follow the intuitive leap that you have made here.  Are you 
>>suggesting that if something is intelligent, it will communicate in a pattern
>>which is like ours?  Are you suggesting that if something communicates in a
>>pattern like ours it is intelligent?  I have put forth the argument that
>>communication is not *necesarily* a reflection of intelligence.  It may well
>>be a reflection of *knowledge* which databases clearly show us can exist
>>without intelligence.
>
>I'm suggesting that the only possible measure of intelligence (until
                                                                ^^^^^
>we define intelligence) is an aggregate of measures of communication.
 ^^^^^^^^^^^^^^^^^^^^^^
>These measures will apply to language (a subject in itself), action 
>(a whole infinite set of subjects (at least until we find a TOE)), 
>reception (which at this point is as undefinable as intelligence), and
>who knows what other 'properties.'  
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

NOT A FLAME: Did you read the subject line?  I argued that the Turing
Test "defines intelligence out of existence".  How can you claim to measure
something which you don't know anything about?  Suppose that I argued that
you could measure weight by measuring volume and then comparing the volume
to a similar volume of water?  That "measuring technique" is valid for things
which weigh the same as water, but is hopeless for iron and feathers -- both
of which have weight.  

You can not set up a valid test for something until you understand what you
are testing!

>
>Without defining those three 'properties' of communication I listed just
>above, I'd like to ask you why communication is not *nec* a reflection
>of intelligence?

I argue that intelligence is an ability.  Communication can be used to
evaluate intelligence.  But then the evaluation is only as valid as the
means of communication (i.e. Poor results may mean poor communication, not
low intelligence).

Brett

bfish@sequent.com



