Re: Platonic Computation and OMs

From: Lee Corbin (lcorbin@rawbw.com)
Date: Sun Apr 06 2008 - 10:07:59 MDT


The post by Stathis

----- Original Message -----
From: "Stathis Papaioannou" <stathisp@gmail.com>
To: <sl4@sl4.org>
Sent: Thursday, April 03, 2008 3:28 AM
Subject: Re: Platonic Computation and OMs

details so many of the key issues surrounding this whole
topic that it warrants re-study. There is material in it that
we can mine for quite a while on this topic, and I think I'll
be returning again and again to it in order to explore
just what is meant by platonic computation.

> Lee wrote:
>> [Stathis wrote]
>> > My mind is apparently closely connected to my
>> > brain, not to patches of dust around the universe,
>> > and I worry a lot more about what happens to
>> > my brain than about what happens to the dust.

It's interesting to note the term "happens" in this sentence.

That term denotes a process requiring temporal duration.
Later (see below) you again emphasize some kind of
temporal connection to the rest of the universe, and so
you seem to be repudiating here what I always thought
of as "the theory of dust" (e.g. taken from Permutation City).

>> > But if the whole universe is actually a simulation
>> > in Platonia, all that means is that the computations
>> > giving rise to my mind through the process laws of
>> > physics -> Big Bang -> stars -> planets -> evolution
>> > of simple life etc. have much higher measure than
>> > those giving rise to a disembodied consciousness.

You evidently do believe that our whole universe (the
level one Tegmark infinite universe) is a "simulation in
Platonia", and the process

physics -> Big Bang -> stars -> planets -> evolution

is much more important in emulating you than would
be non-temporal (timeless) patterns exhibited here
and there randomly in the universe.

To recall cases (B) and (C) again, we have

>> B. by fantastical means collect patches of dust
>> and load them sequentially into memory
>> in such a manner that [the torture seems to
>> be explicitly exhibited and] the sole difference
>> [between that and the usual dreadful events]
>> being that there is no actual causal connection
>> between the states [??]
>>
>> C. just save the time and expense of the loading
>> part of (B) and just leave them laying around
>> in the solar system here and there after we've
>> collected them
>
> The difference between B and C is that B results in a structure
> that interacts with the local universe, like A. That it be able to
> interact with its environment in a meaningful way is the reason
> why the structure of an implementation of a computation is
> important.

I wonder to what extend it truly can be said to be interacting.
My original model, or intuitive example, was of watching an
extremely detailed film, e.g., Casablanca. Far from interacting
with the audience, the portrayed events are merely presented,
even the term "simulation" probably invoking incorrect
associations.

Loading a "succession of frozen" states formulaically---that is,
when the causal connection from one to the next is highly
suspect---strikes me as being "uninvolved" with the environment.
The series of apparent events unfolds quite independently of
the surroundings, (e.g. Turing Test examiners).

It's only when we then introduce via a GLUT the possibility of
of the series of lookups not coming from "the can" (in movie
parlance) but from a huge tree of possible "cans" that concerns
for you first arise that these are not merely "patches of dust",
whose patterns---as you have said---are not important.

> If the structure of a conventional computer is sufficiently
> damaged, then the computation in its useful, interactive,
> recognisable form comes to end.

And so this boils down to more discussion, I guess, about
the GLUT, whereas here I wanted more to get into Platonia,
so to speak.

> But if we are talking about a computation scattered in frames
> throughout time and space, it was never useful, recognisable or
> interactive to begin with; if it computes a consciousness, that
> consciousness will be in an inputless virtual world. And since this
> computation was never going to interact with its environment anyway,
> nothing is lost by disrupting the frames any which way. A hypothetical
> observer could look at any patch of dust and say that he can see how,
> under a certain complicated transformation, it is isomorphic to any
> given computation.

Yes, and then to both you and me, evidently, it would
be meaningless, or (as Bostrom apparently puts it),
without moral or ethical implications, and certainly,
so far as you and I and the other readers here are
concerned, without invoking anticipatory dread.

> Given any apparently random structure or process, there
> exists in the possibility space of alien semantics a printout
> which gives it any particular meaning. Now, while it is true
> that if we had the appropriate printout, the vibrations of the
> atoms in a rock could be seen to spell out the King James
> Bible, or simulate tomorrow's weather in Tokyo, it is only
> of trivial interest, because we don't have that printout.

And it wouldn't matter if we did happen to? After all,
that would only represent some work on the part of
a probably long-ago dead Alien civilization which had
read meaning into a process where it would have been
possible to read any meaning into it whatsoever. Oh,
but then perhaps I have got you wrong:

> But if we consider that the rock is implementing
> computations which lead to conscious observers,
> this is of greater interest; for there is no reason
> why the consciousness of the observers should
> be contingent on whether we have the printout
> or not.

Yes, but how likely do you consider it to be that
any given rock (for the sake of concreteness,
weighing on Earth about the same as a human
being) is actually implementing conscious beings
with measure that is at least a significant fraction
of the computation that your body or mine is
implementing to give us experience?

And following up that question, surely the entire
mass of the Earth's iron core must implement
you or me to such a tremendous extent that our
present metabolisms are relatively of almost
zero significance?

> Thus, all possible computations are implemented
> by mapping onto any physical process. The
> structure is mostly contained in the mapping,
> but in the interesting case of computations
> containing observers, it is sufficient that the
> mapping merely exist in the abstract, rather
> than as a hardcopy or even in the mind of an
> external observer.

If it is sufficient that these mappings exist in the
abstract---which as a mathematical platonist
I surely do agree with you that they do---then
again, I don't quite follow what the significance
is of the fate of the particular Stathis organism
as it presently runs in Australia is all about.

Lee



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:02 MDT