From newshub.ccs.yorku.ca!ists!torn.onet.on.ca!utgpu!news-server.csri.toronto.edu!rutgers!gatech!ncar!noao!amethyst!organpipe.uug.arizona.edu!organpipe.uug.arizona.edu!bill Mon Jun 15 16:04:44 EDT 1992
Article 6213 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!torn.onet.on.ca!utgpu!news-server.csri.toronto.edu!rutgers!gatech!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: The Turing Test is not a Trick
Summary: Aim low, then scale up
Message-ID: <BILL.92Jun11114119@ca3.nsma.arizona.edu>
Date: 11 Jun 92 18:41:19 GMT
References: <1992Jun11.154029.29686@Princeton.EDU>
Sender: news@organpipe.uug.arizona.edu
Organization: ARL Division of Neural Systems, Memory and Aging, University of
	Arizona
Lines: 21
In-Reply-To: harnad@phoenix.Princeton.EDU's message of 11 Jun 92 15: 40:29 GMT

harnad@phoenix.Princeton.EDU (Stevan Harnad) writes:

   Summary: Aim up, then scale down

Here's the thing: the TTT is unpassable now and for the forseeable
future, and so is the TT, given a sufficiently clever examiner.  On
the other hand, machines that can *think*, in the sense of being able
to carry on a broadly-ranging intelligent natural language
conversation, will probably exist in twenty years or so.  I just don't
believe it's reasonable to deny mind-bearing status to such a machine
because it doesn't know that eating sweet things sometimes makes your
teeth hurt.

   First make something that can do everything a car can do and then worry about
   scaling it down to something that can can't drive in reverse or exceed
   0.5 mph. 

As an engineering strategy, this doesn't work too well.  (But :-) it does
have one advantage:  it provides a purpose for the existence of heaven.)

	-- Bill


