Am 31.10.2014 um 20:03 schrieb Richard Fateman:
My advice is you should never get into a situation where a limitation of a
particular
computer program --- or even several computer programs -- leads you to
say that computers cannot compute something "the way humans can".
Recall that for many years people claimed that computers could never
play a good game of chess.
I think the gist of the question was what the computer "does", not what
it "can".
Also, I think it's really a spectrum, and you need to determine where on
that spectrum you wish to go for some particular task:
1) Mimick what a human would do. For some (to-be-determined) definition
of "mimick" and "what a human would do", so this is very vague anyway.
2) Do something wildly different than what a human would do, but
transform the results to something that a human would arrive at
eventually. This is what the vast majority of chess programs tries to do.
3) Do something wildly different than wat a human would do, and
(hopefully) be better than a human's result by some (to-be-determined)
metric.
In the end I think the question whether the computer can arrive at human
capabilities is too narrowly defined - the classical trap of AI: trying
to mimick the human mind instead of exploiting the computer's specific
power will just give us a bad substitute for a human (though research in
that area can still give interesting results).
--
You received this message because you are subscribed to the Google Groups
"sympy" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sympy.
To view this discussion on the web visit
https://groups.google.com/d/msgid/sympy/54540505.3090402%40durchholz.org.
For more options, visit https://groups.google.com/d/optout.
