blow away?
Dan Schwartz <schwartz@iota.cs.fsu.edu>
Date: Mon, 13 Sep 93 23:55:56 -0400
From: Dan Schwartz <schwartz@iota.cs.fsu.edu>
Message-id: <9309140355.AA02198@iota.cs.fsu.edu>
To: interlingua@ISI.EDU
Subject: blow away?
John's statement that AI would ``dry up and blow away'' without
standards raised a red flag for me as well.
A definition I sometimes give to my students is: AI is that field of
endeavor which is concerned with understanding the activities of the
human mind and simulating those activities on a computer. This must
immediately be qualified, however, in light of the fact that humans
frequently make mistakes. Thus one might more correctly say that AI is
concerned with identifying those mental activities which are desirable,
and simulating these on a computer. The practical objective is to use
the computer as an extension of the human intellect, to replicate our
thinking processes for purposes of assisting us in our mental work.
Granted such a definition, then it is almost certain that AI is here to
stay for a very long time. Firstly, we have barely begun to comprehend
the nature of our own minds; new aspects of human intelligence are being
uncovered almost daily by a diversity of academic disciplines. Only
recently, for example, have we begun to appreciate the nature of human
expertise, let alone to fully understand it.
Secondly, human consciousness is itself in a constant state of
evolution, with new capabilities being acquired in a rather
unpredictable manner. An example here is the relatively recent
emergence of psychological techniques for alleviating stress, signifying
a new general level of awareness of basic human emotions. Taken
together, these two observations show that the quest embodied by ``AI''
is totally open-ended, and that it is likely to stay that way virtually
forever.
As to ``standards,'' it seems reasonable that as we identify different
aspects of human mental processes, successfully simulate them, and then
find applications for our simulations, we should then seek to
standardize our methods for purposes of certain application domains.
For example, it makes sense to propose that KIF or Conceptual Graphs may
become standard formulations of first-order reasoning for use in Data
Bases. However, to propose that AI in the broader sense described
above, needs---or even would benefit from---an established set of
representation standards, seems rather far fetched. Any particular
representation scheme is almost certainly destined to be inadequate
except for very limited and well-defined tasks.
--Dan Schwartz
************************************************************************
Daniel G. Schwartz Office 904-644-5875
Dept. of Computer Science, MC 4019 CS Dept 904-644-2296
Florida State University Fax 904-644-0058
Tallahassee, FL 32306-4019 Home 904-385-7735
U.S.A. schwartz@cs.fsu.edu
************************************************************************