From: CurtAdams@aol.com
Date: Mon Jun 17 2002 - 20:20:01 MDT
In a message dated 6/17/02 18:46:31, oregan.emlyn@healthsolve.com.au writes:
>...but way more than smart enough to enhance its own intelligence, one
>would imagine.
In general, the possible interactions (and potential deleterious consequences)
of a change to an intelligence (or any system) should grow faster than the
intelligence
does. It's the old N vs. 2^N Cantor business. Not that a superintelligence
couldn't
become smarter, but there's no reason it's going to go much faster than we are
today with Moore's law. There necessarily remains an enormous gap (us vs.
the sun
kind of thing) between an unquestionable superintelligence and something able
to sim
the planet, and no reason to expect the gap to be bridged quickly.
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:14:52 MST