From: Robert J. Bradbury (bradbury@aeiveos.com)
Date: Tue Aug 05 2003 - 15:18:34 MDT
On Tue, 5 Aug 2003, Robbie Lindauer wrote:
> > Quite right. Robert and many others have made it clear
> > that that's impossible. Besides, given how slow light
> > travels, I have always thought that an integrated intelligence
> > probably can't be more than kilometers in size anyway
> > (depends on the kind of problems that it enjoys and the
> > gratification algorithms it executes).
It is correct that it depends a *lot* on the definition of
"integrated" and a lot on how much time you have to think
about a problem. If one has a long time to think about
problems -- then distributing parts of ones mind across as
solar system (say 10-50 AU) may not be a huge sacrifice
while greatly increasing ones safety. The capacity of
a 1 cm^3 nanocomputer is roughly 100,000 human minds
in something like 1/1000 the volume (of a brain). So
I tend to use 10^6 increase in intelligence as a
conservative approximation. The communications
bandwidth to a copy 25 AU away *will* be low but it
*will* be "you" (or a subcomponent of "you") and so there
will be a lot of shorthand notation. So I suspect there
will be the development of "distributed" computational
algorithms. Examples of this might be the SETI@Home or
the Folding@Home distributed computations of our time.
But you take at least a 3 order of magnitude hit in jumping
from light-minutes/hours to light-years. I just don't think
there are going to be good reasons for doing so.
> Wouldn't it be possible to create a quantum computer based on
> spin-coupling as a communication "backplane"?
Yep, you could propose such a computer. But Robert will jump
all over your butt for proposing "magic physics" on the ExI list.
There are a couple of people who have "Get out of jail free"
cards to do this (Anders & Robin come to mind since they know
a lot more physics than I do). But anyone else gets jumped
on unless it is a serious "what if" kind of speculation.
(The primary objection at least in my mind currently is that
for QC to be useful you are going to have to propose a method
for generating and maintaining kilo, mega, and giga amounts
of qubits. If there isn't a concrete proposal for something
we can actually build, then one gets a "Go directly to jail,
do not pass Go and do not collect 200 dollars" card.
[People who have not played Monopoly will not understand the
references. Sorry...]
Robert
This archive was generated by hypermail 2.1.5 : Tue Aug 05 2003 - 15:28:16 MDT