From: mike99 (mike99@lascruces.com)
Date: Sat Apr 27 2002 - 11:23:26 MDT
Right, Evan, that is the issue. Even the governing AI's that are the SysOp's
for the upload "polis" domains in DIASPORA are not qualitatively different
in intelligence from the upload-minds inhabiting those polises (at least, so
far as I can tell).
If Egan believes what his characters say in his latest book (as quoted by
Damien B.) then apparently Egan views general intelligence as a process that
can be speeded up and given more data processing resources without becoming
qualitatively different. In this view, there will *not* be any emergent
phenomena from faster thought processes that have denser data resources
which could accurately by termed "superintelligence."
No superintelligence, no Singularity.
I certainly disagree with that view. It seems to me that Nature shows many
examples of complex systems exhibiting new behaviors that one would not
expect by a simple analysis of their component parts/subsystems. I believe
this will probably also be true of superhuman levels of intelligence.
Regards,
Michael LaTorra
-----Original Message-----
From: owner-sl4@sysopmind.com [mailto:owner-sl4@sysopmind.com]On Behalf
Of Evan Reese
Sent: Saturday, April 27, 2002 6:35 AM
To: sl4@sysopmind.com
Subject: Re: Egan dismisses Singularity, maybe
----- Original Message -----
From: "Randall Randall" <randall@randallsquared.com>
Subject: Re: Egan dismisses Singularity, maybe
> mike99 wrote:
>
> > A human IQ equivalent of, say, 200
> > may be great and wonderful. But where are the IQ 2000 intellects? Does
Egan
> > believe that a) such a level is unattainable, or b) faster thinking
equals
> > greater intelligence
> Even if he believes "a" or "b", there's also c) it's impossible to
describe
> in detail the actions of someone literally ten or twenty times more
> intelligent than yourself. Such entities cannot really come onstage in a
> work of fiction.
That's true; but he doesn't even refer to them as being off stage. I think
that's the issue.
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:38 MDT