On 09 Sep 2017, at 01:23, Russell Standish wrote:
On Fri, Sep 08, 2017 at 09:48:10AM +0200, Bruno Marchal wrote:
That is right, but fortunately, a computation, when executed, is not
a pile of states, is more like a precisely structured set of states.
We still cannot found the observer there, except for some of them,
but that is not important, because the observer itself can do that.
Bitstring are not enough, here I agree with you.
You have to keep in mind that my theory is a model - the bitstrings
are necessary, but not necessarily sufficient. They represent the data
interpreted by an observer. Something like a universal dovetailer
gives us the bitstrings by virtue of the Washington-Moscow thought
experiment.
?
The iterated self-duplication gives all bitstrings.
The universal dovetailer, or the sigma_1 completeness, by the non
triviality of what is a universal machine/number and what they can
prove and bet on, we better hope not getting all bitstrings, which
with mechanism with white noise.
My argument is that since the only thing we can discuss is appearances
(basically phenomenal physics), and appearances are observational
interpretation of the data, then taking an ensemble of all bitstrings
suffices for working out all that appears in a variety of ensemble
theories.
It is the real number. The number line. Its first order theory is not
Turing universal, and is indeed a decidable theory.
With the function sine, or the complex exponential, you get Turing
universality again.
With mechanism, you can predict that a random oracle is called for
when you observe "reality" below your substitution level, as you are
confronted to the blurry sum on all computational histories going
through your relative actual indexical state.
I well concede that a collection of bitstrings may not be sufficient
to
explain consciousness itself. We're a long way from knowing what might
be sufficient.
I think a universal machinery is sufficient. That one reflect the
whole "competition" in play to bring our continuations.
The elementary arithmetic is full of dreams, and some can cohere and
be sharable in deep "many-users" video game. We are not in "one"
simulation or emulation, we are, in the first person perspective
(singular and plural) in infinity of simulation/emulation.
I put the "big" things like the induction axioms, the infinite sets,
the reals, analysis, physics and (full higher order) theology in the
phenomenology.
The negative numbers, the rationals, the reals, the complex numbers,
the octonions, and the ordinals, and the cardinals, are the
(necessary) imaginary friends of the universal natural number.
We could start from Z, 0, 1, +, *, note. Once you have universal
machinery, the physical reality are like the border of river, except
here the water is played by the flux of first person experience.
There are Turing universal groups (like the unitary group) but we
cant' postulate them, we have to extract them from the machine
introspection if we want exploit the G/G* to keep into account between
the rationally believable, the knowable, the observable, the sensible.
It is enough to use the fact that elementary arithmetic is a "great
programmer". If someone believe that 2+2=4 independently of himself
or of a universe, then, the whole dovetailing is there too. But with
the reals or the bitstring, we get too much things, without enough
structure. You (Russell) are right that the observer recognize its
own computation(s), but you still need the computations for this.
It is fairly uncontroversial to assume that universal computation is
necessary for consciousness, since we humans are capable of that. But
it may not be sufficient.
That is what I thought sometime ago. I thought that full consciousness
require turing completeness + the provable turing completeness (by the
entity in question). But I begin to think that somehow I was wrong.
Consciousness is in turing universality already, I guess as a highly
dissociative state, and from there, the axiom of induction does not
apply.
We have zilch evidence of the
latter. Computationalism is the position that it is both necessary and
sufficient, of course.
I am not sure. If computationalism is true, it can be false too. We
can test it, modulo "bad luck oracle" or "Bostrom type of emulation,
by "malevolent" normal entities. We cannot be emulated in one
computer, as once we are emulated at the correct level, we are in all
emulations. (By the first person indeterminacy on all emulation (whose
existence is a theorem already in Robinson arithmetic.
Computationalism assume a form of self-consistency which cannot be
proved, and in particular, we can't known our substitution level (but
we can bet on one).
I don't know if computationalism is true. I do think the evidences (by
Gödel, and QM) are in its favor.
The great Goddess created 0 and its transitive successors.
Then She said add yourself.
Then She said multiply yourself.
Then She said "oops!".
And lost herself innumerably in the labyrinth of the Dreams emulated
infinitely often in the additive+multiplicative arithmetical reality.
Bruno
--
----------------------------------------------------------------------------
Dr Russell Standish Phone 0425 253119 (mobile)
Principal, High Performance Coders
Visiting Senior Research Fellow [email protected]
Economics, Kingston University http://www.hpcoders.com.au
----------------------------------------------------------------------------
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it,
send an email to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
http://iridia.ulb.ac.be/~marchal/
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
