nazgul@apollo.UUCP (05/24/84)
Be sure to leave a blank line after the header.
Two comments, as follows:
1) I tend to disagree with any statement which contains an absolute
(eg. "computers will never"). However, ignoring that...
2) I tend to agree with the statement that someone (my news-dump is
at home, so I'm not sure who) raised regarding perception over
time vs. perception over space. Namely, there really isn't any
difference. From a philosophical point of view you can continue
to argue about essence and primordial types, but current AI
techniques are still back in the pattern matching stage.
From a pattern matching point of view, "two" objects are actually
the same object iff they match a certain number of criteria. These
generally take into account the number of similarities as opposed
to the number of differences. Additionally certain assumptions are
made if the "two" objects are being viewed at the same time or
at different times. The computer's matching scheme will be just
as capable at doing comparisons of objects seperated by time as of
those seperated by space. (If space and time are simply two different
facets of the same thing this makes sense in a physics context as
well). Certainly it will not always be right, but then, neither are we.
Witness the times you speak to someone you had talked to earilier in
the day and then discover you are actually talking to a twin.
Or the times when you totally fail to recognize someone you haven't
seen in a long time. Matching over time can certainly be improved
by taking into account aging and expected changes, but to a great
degree the only difference that time creates is that our assumptions
are different, the actual matching is the same.
-kee
...decvax!wivax!apollo!nazgul
P.S. Pardon my cloudy thinking, I don't think I've woken up yet
this month.