X-Message-Number: 8175
From: 
Date: Sun, 4 May 1997 20:18:17 -0400 (EDT)
Subject: continuing

1. Metzger had asked me repeatedly to admit I couldn't prove my body is not
in a vat being fed virtual reality. I answered that I had clearly all along
agreed, for purposes of discussion, that we all might be living in a
simulation. This OBVIOUSLY included, as a less far-fetched case, the
possibility he mentions, that I am real but the outside world I think I
experience is virtual. Nevertheless, he again complains (#8166) that he is
talking about the vat problem, not the full sim problem. 

Aside from his failure to see the obvious, what is his point? If I can't
prove I'm not in a vat, so what? *How does this bear on anything at issue?*

2. Again on the computer-freeze problem in the case of simulations generating
different sub-simulations. [If there is one real world with (say) several
real computers each generating a simulation of a world full of people, then
we would expect each of the simulations to generate several different
subsimulations, etc. ad infinitum. I maintain that each of the original
computers in the real world would then effectively freeze up or slow down to
a negligible speed, relative to demand, and in terms of its overall
operation.]

Metzger (#8166) has several missing-the-mark replies:

a. "If you don't believe one machine can efficiently simulate another....go
buy SoftWindows..."  

I never said one machine cannot efficiently simulate another; I said one
machine can't sustain, relative to demand, an infinite cascade of simulations
and subsimulations etc.

b. Metzger: [The subsim-cascade-freeze wouldn't happen because] "...if we are
dealing with a sufficiently detailed simulation, the simulation wouldn't care
if it's simulating a waterfall or a microprocessor." 

The "simulation" is a whole virtual world. The simulated computer(s) in it
behave very differently in certain respects, depending on whether or not they
generate sub-simulations and further sub-subs ad infinitum. The latter case
imposes a RAPIDLY INCREASING LOAD and one WITHOUT BOUND on any level of
computers, and finally on the one or several real computers in the one real
world. 

c. "...you can't build an infinite number of computers in the universe around
you..."

Not germane.

d. To say the real machine will essentially halt "...is a handwave." because
slowing and halting are different.

John Clark (#8168) also points out that, regardless of any degree of slowing
or even occasional halting of portions of the programs, any one of the
simulated people could live what seems to him a normal life, since he
wouldn't notice the time scale changes or intermittent activity. (This is a
bit remindful of the black-hole and similar situations, where outside and
inside time can differ by any amount, without bound.)

What we are dealing with here is something different, however. The postulated
subsimulations pile up with explosive rapidity; there is no convergent
series, so to speak. I call that a freeze.

John Clark also says that simulating an orderly object like a computer might
be easier than a chaotic object like a rock. But we are not talking about
simulating a computer sitting on a shelf; we are talking about simulating a
computer doing an infinite cascade of subsimulations. 

e. [I said the postulated situation would result in zillions of subsimulated
 worlds.]

Metzger: "There isn't space on the planet...for zillions of computers, so why
should a simulation be defective if it doesn't simulate a large enough
universe for 'zillions' of computers?"

Missed the bull's eye, the target, and the side of the barn. I didn't say
anything about "defective" simulations. I said the postulated situation would
demand a cascade of subsimulations, and the real or first-level computer
couldn't keep up with the demand (nor could any of the sublevel computers). 

f. [I previously said that someone who seriously entertained the material
probability that he is living in a simulation ought to be "praying"to his
programmer. Metzger asked "Why?" I said this ought to be obvious.]

Metzger: "...not obvious to the rest of us."

I'm not sure what makes him think he can speak for everybody else on the
list, but if it really has to be spelled out I'll spell it out: 

If I think there is a serious possibility that I am a simulation, then I must
realize that the programmer is capable of knowing what I do and even think,
since all of that can be displayed to him. He also has the ability, from my
point of view, to work miracles-- change history, change my person and
environment, even change the laws of nature (rules of the program).
Therefore, by persuasion or entreaty, I might be able to invoke his powers to
my benefit. This is "prayer."

Metzger goes on to say it would be silly for simulations to pray in this
manner, because the programmer wouldn't be listening and isn't omniscient.

Nobody is talking about omniscience, just about a possible interest of the
programmer in his work. True, he might not pay attention, but again he might.
I would, if I were the programmer. The reward/risk ratio is enormous. In
fact, it is a little bit reminiscent of the reward/risk ratio for cryonics.
No guarantee of success, nor even that you will enjoy success if it comes,
but still the most potentially profitable game in town. 

3. "Sentience" (#8170) mentions possible tests of whether we are living in a
simulation. 

Metzger and others seem to think the sim/nonsim question is impossible to
investigate from the inside. Actually, there are several ways, in addition to
"prayer," that one might obtain evidence on this question, if not 100% proof.
In particular, we might adduce evidence that we ARE in a simulation; evidence
that we are not would be harder.

For example, if continued scientific experiments in physics never produce
surprises--results not implied by previous information--this would suggest
that we are in a simulation, since in the real world we surely anticipate
surprises. (Of course, this would not be absolute proof; there could be
programmed "surprises.")

As for "Am I in a vat?"--this also might, in some circumstances, be amenable
to investigation by the subject. For example, if I suspect a certain enemy
has done this to me, then my body is in some specific location, and I can try
to find it. Depending on the nature of the VR feed, I do have some
interaction with the outside world, and if this is close enough to normal I
might even find my body, using my virtual senses.

More generally, I suspect there must ALWAYS be bidirectional interaction
between worlds (a kind of generalized Newton's Third Law), so that "escape"
in some degree might always remain possible, however difficult.

Robert Ettinger

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=8175