X-Message-Number: 8185
From: 
Date: Tue, 6 May 1997 21:59:08 -0400 (EDT)
Subject: misunderstanding Zeno

First, to those who think these exchanges time-wasting for the reader, you're
not compelled to read them. To those who think the writers could make better
use of their time, please be assured that this stuff is a very small part of
my activity, which mainly involves mainstream cryonics work including trying
to coordinate several lines of research. It will also have some use in one of
the books I am working on. 
-------------
Metzger's # 8180 continues his pathological habits of evasion, red herrings,
subject changes, lexicon changes, and the use of straw men--attributing to me
positions I have not taken nor implied. This wrangle is distasteful, but I
won't give the impression of conceding, so I'll go on. (Those who know me
also know that, when I am wrong, I am not at all reluctant to concede.)

The main claim in Metzger's #8180 was that I started the discussion by saying
some one couldn't live out his life in a simulation, and that I have now
conceded that I might be doing so.

Wrong as usual.

What I have said, among other things:

a) We do not know that an inorganic medium--let alone a computer
simulation--could sustain feeling and consciousness. The "information
paradigm" is only a postulate, not proven.

b) IF a simulation could be conscious, and IF a computer could be made to
simulate a world full of people, still generally one could not live
indefinitely, in such a simulation, a life closely similar to what one's life
would have been in the real world. This is because--among other things--the
"laws" of the simulation could not anticipate discoveries of new laws in the
real world, so that a simulated scientist, for example, could not
experimentally discover phenomena not implied by previously known rules. 

c) If people in the real world are motivated to simulate a world full of
people, the simulated people will also be so motivated, generating an
exploding set of subsimulations, effectively causing the original computer,
in the real world, to stop. 

Now some more specifics pertaining to Metzger's # 8180:

1. Metzger: "...you [contended] ...that someone couldn't live out their life
inside a simulation......now you are saying...you might be in a
simulation....a slight contradiction?"

No, no contradiction. It may be a logical possibility that my natural body is
in a vat experiencing virtual reality. That has nothing to do with the
question of whether a simulated person in a simulated world could live out
his life in the same manner (with the same events) as in the real world. He
could not, one reason being that the laws of nature are not fully known,
hence cannot be embodied in a simulation.

2. I had said that the rapidly branching subsimulations would produce an
unbounded, rapidly increasing load on every level of computers (one or two or
a few real and the rest simulated).

Metzger:  "...false...the total amount of computation being done by the
system is bounded, because the sum of any geometric decreasing series is
finite." 

No, he has it wrong on two counts. First, I didn't say the the amount of
computation done by the system is unbounded, but that the amount being ASKED
(the offered "load") is unbounded. Second, the computation actually being
done by the one real computer (or more than one) is obviously bounded in any
bounded time interval--not because of any geometric series, but just because
a real computer simply chugs along at its own rate regardless. 

Metzger tries to defend his position by reference to one of Zeno's
paradoxes--Achilles and the Tortoise, which is explained in terms of a
convergent series, as most readers here probably know. But we are NOT dealing
here with a convergent series in the sense he means. A single, limited
computer is being "asked" to generate an unlimited number of subsimulations
(all a little different) in a limited time, and it can't. You don't need much
math education to understand that.

The REAL convergent series here is one which DISPROVES Metzger's claim. As
the number of subsimulations explodes, the amount of real computer time
available to each subsimulation, per unit of subsimulation time or per subsim
state transition, rapidly approaches zero. Depending on the details, then,
one could calculate a stopping time as viewed from any of the subsimulations.

The result could of course depend on the structuring of the program; the
machine might e.g. be told to allocate its time so as to focus on some
aspects and ignore others, as John Clark said. But in any case it CANNOT do
the job it is asked to do in our postulated experiment. 

Again: my use of the word "load" did not imply that the real computer would
belch smoke under the stress, or that the simulated computers would belch
simulated smoke. It obviously just meant that it couldn't do what was asked,
and would effectively stop. The real computer keeps ticking at a zillion MHz
or whatever, but the results it produces go nowhere (in any given
subsimulation).

3. Metzger: "You claim it is a flaw of the simulation that it cannot sustain
an infinite amount of computation."

For what I said, see above. Find another wind mill or straw man.

4. Metzger: "I am currently running five nested copies of PSIM on the machine
I am typing this on. The outer machine is not "grinding to a halt."

I did not say the hardware of the real computer grinds to a halt; the
effective production of the system grinds to a halt. See above. And instead
of five, try running a billion nested copies on your machine--just for
openers.

5. Continuing the "prayer" argument, Metzger disregards my main points and
offers possible reasons why "prayer" might not work.  He even talks about
"creating a religion," when my remarks clearly had no such suggestion. (If
you logically conclude that there may be a Programmer out there, and that you
might be able to influence him/her by thought or talk, that isn't religion as
usually conceived.) 

He disregards all probability calculations and thinks options should be
dismissed on the basis of possible problems, without assessing degrees of
likelihood. He also says we "know nothing" about the putative programmer; but
if we are hypothesizing that we are SUBsimulations, then the programmers are
probably people much like us. Is that so hard to understand? 

Once more, I personally don't take the possibility seriously, because I think
it highly unlikely that a simulation could be conscious; but for someone with
the info viewpoint it is a different story.

6. As to the vat problem, I said that--depending on details of the setup--I
might be able to investigate the real world, even locate my body. Metzger
says the brain in the vat, hooked up to a simulation (virtual world) would
have no interaction with the outside world.

As I said, it depends. By hypothesis, my normal brain (perhaps while
anaesthetized) has been disconnected from the rest of my body and connected
to a virtual world. But unless my memories, personality etc are totally
reworked, the signals I get must seem normal, and what happens when I make
virtual movements, as going for a virtual walk, must seem normal. If this
"normality" is provided by using the real world for input, then I might
indeed, for example, go to the library (virtual library just like the real
one) and read a book (virtual book just like a real one); I might also go
(virtually) to where my vat is located and (virtually) see it......None of
this is important or plausible; it is only intended to show that things that
might initially seem impossible are not necessarily so.  Again: it is
entirely possible that the brain in the vat might be able to adduce evidence
that it is or is not.

In any case, the possibility or difficulty of getting proof or evidence for
the vat says nothing at all about the claim that someone could live out his
life as a simulation.

Robert Ettinger

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=8185