On 13 Aug 2014, at 22:29, meekerdb wrote:
On 8/13/2014 7:16 AM, Bruno Marchal wrote:
On 12 Aug 2014, at 18:54, meekerdb wrote:
On 8/12/2014 6:36 AM, Telmo Menezes wrote:
On Tue, Aug 12, 2014 at 6:17 AM, meekerdb <[email protected]>
wrote:
On 8/11/2014 9:27 PM, LizR wrote:
On 12 August 2014 15:50, meekerdb <[email protected]> wrote:
On 8/11/2014 8:26 PM, LizR wrote:
Well, I guess a physical UD would be made robust against
quantum uncertainty, like all computers, but why do we need to
assume QM apply?
The argument assumes it doesn't apply, so that the computation
can be deterministic. I don't know that it affects the argument,
but worries me a little that we make this unrealistic
assumption; especially if we have include a whole 'world
context' for the MG simulation.
Ah, I see. One of the assumptions of comp is that consciousness
is a classical computation. At least I think that's what it
means to say that the Church-Turing thesis applies. I suppose a
question here is whether QM can introduce some "magic" that
allows it to create consciousness from a purely materialistic
basis. If so then there's no need for comp because consciousness
isn't classically emulable....yes?
Although a quantum computer can compute some things much faster
than a classical computer, it still can't compute things that are
Turing uncomputable, so I don't think it provides that kind of
magic. I was thinking more of the fact that the recorded inputs
to B and the response to the projection of the movie onto the
graph will not be perfectly deterministic, but only with high
statistical probability. Also, in QM it generally makes a
difference to the evolution of the system
whether other states are available even if they are never occupied.
This is what I don't see. Why do A's internal processes have
meaning, while B's don't - given that they're physically
identical?
B's have meaning too, but it is derivative meaning because the
meanings are copies of A's and A's refer to a world. So it's an
unwarranted conclusion to say, see B is conscious and there's no
physics going on. There's plenty of physics going on in the
past that causally connects B to A's experience. Just because
it's not going on at the moment B is supposed to be experiencing
it isn't determinative. Real QM physics can require
counterfactual correctness in the past (e.g. Wheeler's quantum
erasure, Elitzur and Dolev's quantum
liar's paradox).
Well, as I've mentioned previously I think time symmetry may
sort out those awkward retroactive quantum measurements. But
anyway, I guess this is putting the schrodinger's cat before the
horse, in that comp only assumes classical computation and
attempts to derive a quantum world from it. So I guess we can't
necessarily assume real QM physics, or at least not unless we've
shown comp to be based on false premises or internally
inconsistent, or have a rival theory of consciousness arising
naturally from qm and materialism, or some other good reason to
do so. I think what I'm trying to say here is that to assume
comp must work with real physics is to assume from the start
that there is no reversal.
Well there's also the question of whether comp and the UD solve
the hard problem any better than psychophysical parallelism.
I'm not sure that comp and UD propose to solve the hard problem,
as much as proposing why it's not solvable.
I agree. I don't think it's solvable in the way people ask for.
I think it's solvable in the engineering sense. The advantage of
Bruno's theory is that, if the theory is right, then he can prove
within it what the hard problem is not solvable and can say why.
OK. Cool.
Pierz did a good job of examining this and I made some comments
on his post. I would like to look at comp+UD as just another
scientific hypothesis which we will adopt when it makes
some surprising prediction which is proved
out by tests. Obviously getting some surprising, trestable
prediction out of it is likely to be very difficulty. But unlike
Bruno I'm not much persuaded by logical inference from logic,
Church-Turing, or Peano arithmetic because I think they aren't
"The Truth" but just models we use in our thinking. Just reflect
on how all logicians and philosophers would have said, "No object
can be in two different places at the same time. It's just
logic." - before quantum mechanics.
It is my impression that progress in logic is done by removing
all that is non-abstract. It's a simplification effort. My
difficulties with logic usually arise from not being able to
grasp the counter-intuitive simple level at which it operates.
Confusing common sense with logic is a common mistake. You see
this a lot on you tube these days, where well-meaning atheists
like to say "it's just logic" when they are in fact referring to
scientific common sense. I am an atheist, an agnostic and a lover
of science, so I never like this -- it's resorting to the tricks
of the "enemy".
I'm not sure I follow you here. Why does making the simulation
bigger invalidate the argument? Is there a cut-off point?
I don't know about a cut-off. The argument is a reductio. The
conclusion Bruno makes is that no physical process is necessary
to support consciousness,
OK
consciousness can be instantiated in a Turing machine simulation.
Sorry to split the sentence, but I must admit I thought that
latter part was his initial
assumption, rather than his
conclusion?
The initial assumption is consciousness can be instantiated by a
physical computation (one that replicates the I/O of your
neurons), but step 8 is to show it must be independent of the
physical computation and can be instantiated by an abstraction.
But my argument is that the simulation must also simulate a
world that the consciousness interacts with, is conscious *of*,
that a physical world is necessary for consciousness. If it's a
simulated consciousness, then it can be a simulated physics but
it has to be some physics.
Right, yes, I see. Or I think I see. That's implying that the
comp argument is assuming what it sets out to show, that is, it
sets out to show that physics can be derived from consciousness
as computation, but if it has to introduce physics to show this,
then the argument has become circular. So if interactions with
an environment are necessary for consciousness to exist (as part
of the definition of consciousness) then the argument is
necessarily circular. The question is whether the interaction is
necessary, or incidental - "incidental" would mean that
consciousness has arisen in a physical world through evolution,
and hence is highly specialised as an agent interacting with
that world, but it could at least in theory arise some other way
(e.g. inside a computer). Although it's hard to imagine how any
conscious being could learn anything useful without interacting
with some sort of world - it would sure be a blank slate
otherwise. So I guess the question boils down to: is a blank
slate consciousness - one that isn't aware of anything (except
its own existence, I guess) possible?
It is in Bruno's conception. It is MOST conscious because it can
go anywhere from there, be anybody or any being. That's why he
thinks intelligence, which he deprecates as mere "competence",
detracts from consciousness. It has narrowed or directed
consciousness. As you can see that is quite different from my
idea of consciousness as something that arose as a way for
evolution to take advantage of perception mechanisms in doing
learning, prediction, and planning. I think consciousness is a
certain kind of thought and it's about something. Bruno thinks
it's a mystic property of relations between computations, e.g.
being provable.
In a more naive way, I arrived at the same conclusion before
reading Bruno: that intelligence and consciousness are different
things. I did it through introspection. I know highly intelligent
people that think like you. I know I am being honest and I fully
believe you are being honest too, so I see this divergence as
part of the mystery, that merits investigation.
I agree that intelligence and consciousness are different, and I
can see Bruno's point that competence tends to take away awareness.
OK. usually I distinguish competence (which is domain dependent,
and measurable locally) from intelligence. But it is a good idea to
identify consciousness with intelligence, at least before nuances
can impose itself (intelligence might be close to conscience than
consciousness per se, and requires some more self-
referential ability. Non Löbian machine can be conscious, as I tend
to think now that all universal machine are conscious, even if
disconnected from our reality), and so intelligence is closer to
self-consciousness, or Löbianity. Your 3p description obeys the
modesty Löbian formula, that is you believe []p -> p only if you
believe p (formally []([]p -> p) -> []p).
Knowing a lot about a subject keeps you from seeing it with fresh
eyes.
And worst, it can make you believe that you are intelligent, which
is the root of stupidity---almost a comp sin or blaspheme.
But human-like consciousness entails seeing and distinguishing,
it's model making as a way to seeing the world.
OK.
And mathematics, logic, and arithmetic are models too.
Not OK. You confuse reality (model, in the painter or logician
sense) and model in the sense of physicists, which is mainly the
logicians "theories" or "machines".
PA is a "model" (a theory), but the standard model of arithmetic is
the arithmetical reality, or arithmetical truth, and it is provably
beyond any effective theory (effective means that the proofs are
mechanically checkable).
But the arithmetical "reality" only exists in sense of being a model
(in the mathematicians sense) of Peano's axioms or similar and in
particular the axiom of infinity.
Yes, that is coherent with the fact that PA cannot define it (in
particular PA has no infinity axioms). We don't need to formalize the
internal level, but we can, as any Löbian machine can decide at some
point to believe in some axioms of infinity going from being PA to
being PA + <>t or even being ZF.
Platonists just have a prejudice against "material" models - even
though the "material" is Hilbert space or Calabai-Yau manifolds.
Well, with comp, it is no more a prejudice. It is a fact that we
have to explain matter in term of something else.
It's a fact "with comp". In other words, assuming comp is true then
it must follow from comp - which I think is a tautology. But it
comes back to the ambiguity of "comp". Is it just the bet on saying
yes to the docotor, or does it include all that you claim to follow,
including the reversal of step 8?
The definition of the comp theory does not. But the consequence
follows, unless there is a flaw. So comp means just the bet on saying
yes, in september, and it should mean (modulo the possibility of a
flaw) what you get from the UD Argument. Like elementary arithmetic
means the axioms of RA, at the beginning, and the theorem (in RA)
about universal numbers later.
Of course if you explain anything you must explain it in terms of
something else, presumably in terms of something better understood.
UDA use comp, but the result is that the theory of everything needs
only elementary arithmetic. But the Löbian entity, which exists in RA
(but is not RA!) will develop stringer beliefs (set, consciousness,
physical universes, etc.)
If matter is described exclusively by Hilbert space or Calabai-Yau
manifolds, that has to be proven from any Turing universal system
(or its specification in some logic).
What if it's not exclusively? Proofs can only be from axioms in
accord with some rules of inference.
If it is not exclusive, it makes the multiverse bigger, with "region"
not handled by Hilbert space or Calabi-Yau manifolds. Those becomes
geographical reality, contingent one, and no more physical laws (which
by definition of "law" concern the whole physical reality).
Bruno
Bruno
--
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 http://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 http://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
