🚀 go-pugleaf

In article <4251@svin09.info.win.tue.nl>, luit1@wsinfo11.info.win.tue.nl (Wim Jan
sen) writes:

|> erik@maui.til.com (Erik Horstkotte) writes:

|> >As far as whether I regard fuzzy logic as part of AI or not, the
|> >answer is "not really."  The main reason for this is that I'm one
|> >of the dying breed that still thinks that AI is the search for ways
|> >to make computers think *like humans.*  I don't think I've ever
|> >heard anyone seriously claim that humans use fuzzy logic internally,
|> >although fuzzy logic does model certain aspects of human thinking
|> >fairly well.
|>
|> I have to disagree with you here. Although humans do not use fuzzy logic in
|> the process of *thinking*, they do use it in the process of *thinking
|> about*. My point is that you are absolutely correct, if you say that neural
|> networks have not much in common with fuzzy logic. However, fuzzy logic is
|> a very good model for reasoning of humans.
|>  I always though that the main difference between a machine and a human was
|> the accuracy with which they work. Computers calculate with discrete
|> quantities, whereas humans use inaccurate (fuzzy) measurements.


For those interested, there is a nice interview with Prof L.A.Zadeh in which
he talks about AI v fuzzy in detail in

"Fuzzy Sets and Applications: selected papers by L.A.Zadeh" (eds R.R.Yager,
S.Ovchinnikov, R.M.Tong, H.T.Nguyen), John Wiley & Sons, 1987.

He echoes the sentiments expressed above. To quote the beginning of the
interview, "the tools researches use to probe certain AI problems, says
Berkeley professor, are too precise to deal with the "fuzziness" of the real
world."


|> Wim Jansen


--------
P.V.R. Narasimha Rao