From: James Rogers (jamesr@best.com)
Date: Tue Mar 13 2001 - 12:37:31 MST
At 12:03 PM 3/13/2001 -0700, Brent Allsop wrote:
> I'm looking forward to telling you all "I told you so" when
>they finally discover qualia in a few years, start effing ("Oh THAT'S
>what salt tastes like")... and prove that you can't just simulate
>consciousness on any old only causal machine.
I don't see how this makes a difference; do you define consciousness as
somehow being fundamentally different than intelligence? If we can produce
an intelligent machine, I don't think it really matters whether or not it
knows what salt tastes like. It is equally useful as an AI whether it
fails your definition of consciousness or not; I certainly don't think
there is any real evidence that qualia is required for intelligence. I
don't care if an AI experiences red like I do (and we can't really know how
it experiences it at all), but if it can identify red (trivial) then it has
every bit as much utility as a human with regards to "redness".
> Red will always be fundamentally and primally red, and nothing
>else will ever be "like" it. To so completely ignore this fact or
>even its possibility is pure philosophical idiocy if you ask me.
So? I don't grasp the importance. Interesting perhaps, but I can't see it
as anything more than a secondary effect. It seems to me that you are
putting undue weight on a relatively inconsequential phenomenon. I know
this has been gone over many times on the list, but this type of emphasis
on qualia always comes across as a rather arbitrary fixation. Maybe I'm
just uneducable, but no one has ever said anything that has made me put any
real value on it.
>Let's not miss this opportunity to
>anticipate THE greatest of all scientific descoveries ever made - and
>that will be the discovery of the phenomenal qualities of conscious
>qualia.
I fully expect the basis for qualia to be not very interesting in the end;
a one horse town off the beaten track to artificial intelligence. I don't
want to say "Who cares?" because some people obviously do, but even if what
you say is largely true, what is the supposed benefit that plain old FS
machine intelligence won't get us?
-James Rogers
jamesr@best.com
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 08:06:20 MST