From: Mark Walker (mdwalker@quickclic.net)
Date: Sat Dec 14 2002 - 08:32:03 MST
----- Original Message -----
From: "Max M" <maxmcorp@worldonline.dk>
> I have noticed that this is a normal viewpoint for tranhumans. I have it
> myself to some degree. But after thinking about it, it has really
> started driving me up the wall.
>
> The argument goes: "After the singularity, we will be so intelligent and
> different that we will be impossible to understand for mere humans".
>
> What I hate about it, is that it is exactly the same kind of mystiscism
> that religions use. It's dogmatic and un-scientific. It cannot be tested
> in any way. It's no different than saying that "after death we will all
> go to heaven." Wich cannot be tested either.
>
> Perhaps we can't understand it, but we should at least try to. We should
> at have some theories that could be tested.
>
I argue in "Naturalism and Skepticism"
http://www.markalanwalker.com/natu.htm that it is possible to test the
conjecture that we can create higher intelligences even if we cannot
understand in detail the cognitive life of higher intelligences. According
to at least some definitions of 'posthuman' or 'higher intelligence', it is
a necessary condition that their cognitive lives transcend our ability to
fully understand. Presumably you have another definition in mind, otherwise
your exhortation to understand such entities makes no sense. What does not
follow is that, if we cannot understand in detail the cognitive lives of
higher intelligences humans, we cannot verify their existence. To think
otherwise is to subscribe to some version of the dogma that if some subject
S verifies the truth of some theory or proposition P, then S must understand
P. In other words the dogma is that , understanding or comprehending P is
necessary for verifying the truth of P. A simple example shows this dogma
false: S is a child aged 7. She asks her physicist mother if the proposition
'E=mcc' is true. Mum says yes. S has verified that P but has very little
compression of what P means. No doubt you can see where this is going. We
may stand in the same relation to higher intelligences. We may be able to
verify that the conjectured higher intelligences are indeed higher
intelligences. Imagine such a being writes down an enormously complex
equation of the universe that we cannot even begin to understand. To show us
some proof that this equation shows a better insight into the nature of
reality the higher intelligence puts on a display by re-arranging the stars
in the night sky to form the outline of the comeback Elvis. Surely as you
stand there in your backyard looking at the outline of Elvis in the night
sky you would have to think there is some reason to believe that you have
witness an impressive display by a higher intelligence. Of course the
discouraging part might come when the higher intelligences say to us that
their most impressive achievements are not appreciable by us anymore than a
chimp can appreciate our intellectual achievements when we pass them copies
of The Critique of Pure Reason, or The Origin of Species.
I agree in general with the idea of testability. It seems extremely
improbable to me, but I think we should bear in mind the empirical
possibility that higher intelligences cannot be created. As an analogy,
there are limits to how high a bird can fly no matter how much we reengineer
them--at some point the air becomes too thin for any amount of flapping to
be useful. Perhaps we stand at the top of what is intellectually achievable
by any species (carbon or silicon). As I said, I think this is extremely
unlikely, but it has to be left open as an empirical possibility. Suppose
anything smarter than a human reaches a "critical mass" of some emergent
property and the intelligence explodes. Is this possible? Yes. Is it
probable? No.
Mark
Mark Walker, PhD
Research Associate, Philosophy, Trinity College, University of Toronto
Editor-in-Chief, Journal of Evolution and Technology www.jetpress.com
Editor-in-Chief, Transhumanity, www.transhumanism.com
www.markalanwalker.com
This archive was generated by hypermail 2.1.5 : Wed Jan 15 2003 - 17:58:44 MST