Re: AI

Dan Schwartz (schwartz@nu.cs.fsu.edu)
Mon, 21 Sep 1998 20:31:46 +0200 (MET DST)

Dear Biscers and others,

I read with interest the recent exchange between Herb Simon and Lotfi Zadeh.
It seems to me that among the "successes" of modern AI, one might put

Introduction of the first intuitively plausible model of natural human
reasoning with imprecise information, a concise statement of the
general problem, and the consequent opening of an entirely new field of
research,

and among its "open questions" one might put

How to create increasingly more comprehensive models of ever more
sophisticated aspects of this sort of reasoning that can be effectively
implemented on modern day computers.

In this manner, the work in fuzzy logic could be embraced as an important
and highly valuable contribution to mainstream AI, rather than a pariah
deserving only of continued derision.

Likewise, on the fuzzy logic side, it would be useful to acknowledge that
symbolic methods do have their role, and may well harbor opportunities for
many future advancements. Surely, neither approach has reached its ultimate
fruition, nor does either hold the final answers.

But strange, isn't it, that after 25 years of relentlessly discounting fuzzy
systems theory as just a lot of rubbish, the mainstream adherents of AI
should bristle with such anger at the slightest hint of criticism of their
own domain. For my part, over the years, I have found such negative
attitudes towards fuzzy logic appalling. It as if the people in AI are
claiming that the many hundreds of scientists around the world who have
dedicated large portions of their careers to this subject are just a bunch
of idiots or fools. Such can only be branded as arrogance, and it is
unbecoming of a discipline supposedly dedicated to the advancement of human
knowledge.

In the future, though, people likely will be primarily concerned only with
benefitting from the successes and contributions of both camps, and will
regard the current controversy as largely irrelevant. I suspect, moreover,
that this battle will be found to have more to do with the allocation of
research dollars than it does with matters of scientific merit. Proponents
of mainstream AI have long been jealous of money drifting away from their
camp and have steadfastly fought to stop this flow. Such is only human
nature. One tends to favor one's own over others, and regrettably, the
people in AI have succeeded in holding more sway with the American granting
agencies than have their fuzzy systems counterparts.

But in the end, truly sound ideas always carry their own weight and
ultimately withstand their own tests of time. Symbolic AI has been around
since the middle 1950's, and fuzzy logic since the middle 1960's, and both
have had their successes, as well as their share of hyperbolic claims.

I propose that the proponents of both camps openly acknowledge each others'
achievements and shortcomings, and move on. The available funding for
theoretical work in either fuzzy logic or symbolic AI is nowadays very slim.
And this is not due to a lack of worthy problems. It would be good for both
sides to join forces and begin a campaign for a new infusion of research
dollars, to renew and revitalize the exploration of foundational issues at
the heart of our discipline. Investigations of machine intelligence have
much to offer the broader realms of computing and information technology,
and this fact needs to be clarified and advertized.

In my opinion, we simply cannot afford to continue working at odds with one
another. In the spirit of "united we stand," I encourage that both sides
try to find a way to resolve their differences and work together towards a
common cause.

--Dan Schwartz

************************************************************************
Daniel G. Schwartz Office 850-644-5875
Dept. of Computer Science CS Dept 850-644-2296
Florida State University Fax 850-644-0058
Tallahassee, FL 32306-4530 schwartz@cs.fsu.edu
U.S.A. http://www.cs.fsu.edu/~schwartz
************************************************************************

############################################################################
This message was posted through the fuzzy mailing list.
(1) To subscribe to this mailing list, send a message body of
"SUB FUZZY-MAIL myFirstName mySurname" to listproc@dbai.tuwien.ac.at
(2) To unsubscribe from this mailing list, send a message body of
"UNSUB FUZZY-MAIL" or "UNSUB FUZZY-MAIL yoursubscription@email.address.com"
to listproc@dbai.tuwien.ac.at
(3) To reach the human who maintains the list, send mail to
fuzzy-owner@dbai.tuwien.ac.at
(4) WWW access and other information on Fuzzy Sets and Logic see
http://www.dbai.tuwien.ac.at/ftp/mlowner/fuzzy-mail.info
(5) WWW archive: http://www.dbai.tuwien.ac.at/marchives/fuzzy-mail/index.html