Le 26-nov.-06, à 07:09, Colin Geoffrey Hales a écrit :
> I know your work is mathematics, not philosophy. Thank goodness! I can
> see
> how your formalism can tell you 'about' a universe. I can see how
> inspection of the mathematics tells a story about the view from within
> and
> without. Hypostatses and all that. I can see how the whole picture is
> constructed of platonic objects interacting according to their innate
> rules.
>
> It is the term 'empirically falsifiable' I have trouble with. For that
> to
> have any meaning at all it must happen in our universe, not the
> universe
> of your formalism.
Let us say that "my formalism" (actually the Universal Machine talk) is
given by the 8 hypostases. Then, just recall that the UDA+Movie Graph
explains why the appearance of the physical world is described by some
of those hypostases (person point of view, pov).
> A belief in its falsifiability of a formalism that does
> not map to anything we can find in our universe is problematic.
The "intelligible matter hypo" *must* map the observations. If not, the
comp formalism remains coherent, but would be falsified.
>
> In the platonic realm of your formalism arithmetical propositions of
> the
> form (A v ~A) happen to be identical to our empirical laws:
>
> "It is an unconditional truth about the natural world that either (A is
> true about the natural world) or (A is not true about the natural
> world)"
Physics does not appear at that level.
>
> (we do the science dance by making sure A is good enough so that the
> NOT
> clause never happens and voila, A is an an empirical 'fact')
>
> Call me thick but I don't understand how this correspondence between
> platonic statements and our empirical method makes comp falsifiable in
> our
> universe. You need to map the platonic formalism to that which drives
> our
> reality and then say something useful we can test.You need to make a
> claim
> that is critically dependent on 'comp' being true.
If comp is true the propositional logic of the certain observable obeys
the logic of the "intelligible matter" hypostases, which are perfectly
well defined and comparable to the empirical quantum logic, for
example.
We can already prove that comp makes that logic non boolean.
>
> I would suggest that claim be about the existence or otherwise of
> phenomenal consciousness PC would be the best bet.
You have not yet convince me that PC can be tested.
>
> There is another more subtle psychological issue in that a belief that
> comp is empirically testable in principle does not entail that acting
> as
> if it were true is valid.
You are right. The contrary is true. We should act as if we were
doubting that comp is true. (Actually comp is special in that regard:
if true we have to doubt it).
Note the "funny situation": in 2445 Mister Alfred accepts an artificial
digital brain (betting on comp). He lives happily (apparently) until
2620 where at least comp is tested in the sense I describe above, and
is refuted (say).
Should we conclude that M. Alfred, from 2445q, is a zombie ?
> Sometimes I think that is what is going on
> around here.
>
> Do you have any suggested areas where comp might be tested and have any
> ideas what the test might entail?
Naive comp predicts that all cup of coffee will turn into white rabbits
or weirder in less than two seconds. Let us look at this cup of coffee
right now. After more than two seconds I see it has not changed into a
white rabbit. Naive comp has been refuted.
Now, computer science gives precise reason to expect that the comp
prediction are more difficult to do, but UDA shows (or should show)
that the whole of physics is derivable from comp (that is all the
"empirical" physical laws---the rest is geography). So testing comp
needs doing two things:
- deriving physics from arithmetic in the way comp predict this must be
done (that is from the pov hypostases)
- comparing with observations.
The interest of comp is that it explains 8 povs, but only some of them
are empirically testable, but then the other appears to be indirectly
testable because all the povs are related.
To sump up quickly: Comp entails the following mystical propositions:
the whole truth is in your head. But I have shown that this entails
that the whole truth is in the "head" of any universal machine. I
explain how to look inside the head of a universal machine and how to
distinguish (in that head) the physical truth (quanta) from other sort
of truth (like qualia). Then you can test comp by comparing the
structure of the quanta you will find in the universal machine "head",
and those you see around you in the "physical universe".
It is not at all different from the usual work by physicists, despite
it makes machine's physics(*) a branch of number theory. We can compare
that "machine's physics" with usual empirical physics and test that
machine's physics.
(*) "machine's physics" really means here the "physics extracted by an
ideally self-observing machine".
>
>>> ... I am more interested in proving scientists aren't/can't be
>>> zombies....that it seems to also challenge computationalism
>>> in a certain sense... this is a byproduct I can't help,
>>> not the central issue.
>>
>>
>> I can have a lot of sympathy for an argument showing that
>> science cannot be done without consciousness (I personaly
>> do believe this!). But if the byproduct is that machine
>> cannot be conscious, then I think it will make your argument
>> far less interesting, and, by Church thesis, necessarily
>> non effective (if it was a machine would be able to produce
>> it). Actually, the acomp hypostases can already been used
>> to explain why machines will develop argument why their
>> own consciousness are special and cannot be attached to
>> any describable machine in a provable way (and they will
>> be 1-correct!!!). It is the diabolical (somehow godelian)
>> prediction of comp: machines will not believe in comp,
>> just bet on it in some circumstance.
>>
>
> Conscious machines will happen. And it will be very very
> interesting...With any luck it'll be me who does it. It's my mission.
> My
> PhD.
I guess you are joking :)
But I wish you success in the PhD.
>
> What I contend is that abstract computation alone will not do it. Part
> of
> the machine physics, apart from being able to shuffle charge around in
> accordance to the symbols, is to ALSO attach the physics of experience.
> (Is this what COMP is pointing to?).
Yes thanks to the fact that all hypostases come in couple (like G, G*;
Z, Z*, X1 X1*, etc.) which gives the sharable and the non sharable
truth. This gives room for consciousness, qualia, PC, ...
> Then the machines will have
> phenomenal consciousuness and as a result be a whole bunch more
> adaptable
> and smart. And get seats on planes.
Recall that if digital machine have no PC, then it means you have to
attach PC to something not turing emulable. Some mathematics can help
to study those non turing emulable entity, but there is not yet any
evidence such a thing exists in"nature".
>
> At that time (a few years.. 10 max?) I'll be able to test comp for
> real in
> teeny weeny scientists. My prediction is comp is false in our universe,
> but true in the platonic realm where existence and computation of
> abstractions are identities.
In that case platonia owns myriad of zombies.
> That is not our universe, IMO, where
> 'existence' is 'computation of/by <blah>' and the 'blah' is not nice
> neat
> mathematical ideals. Depicting it with platonic ideals is very useful
> and
> informative...but that's as far as it goes.
After Godel we know that Platonia is full of typhons, tempest, dusts,
ten thousand living things, war, etc.
>>
>> The distinction between computability and provability is
>> fundamental for the AUDA. Alas, you are not the only one
>> who miss the distinction, but explanation of this has to
>> be a bit more technical. I have already attempted some
>> explanation, but it is perhaps too hard to convey in a
>> non technical list. I don't know, presently.
>
> It is something I am working on. Any light you can shed on it I'd
> greatly
> appreciate.
Very shortly: all attempt to defined the class of all "intuitively"
computable functions has given the same class of functions, and this
despite very different motivations (lambda conversion, rational quantum
mechanics, turing machine, billiard balls, game of life, etc.).
There is a notion of universal machine with respect to computation.
Universal dovetailing makes sense.
For provability, no two theories have the same theorems in general.
There exist transfinite chain of more and more powerful theories, etc.
There is no universal machine with respect to proof. Universal
dovetailing does not make sense.
> I have no clue what the quotes re angels and G etc are about.
Normal. Those things does not make sense if you have not get a good
understanding of the difference between provability and computability.
> However,
> once you can accept 'wild-type' naturally occuring primitive "objects"
> as
> available (as opposed to idealised platonic realm primitives like
> integers
> with nice neat relationships) then their inter-relatedness is a formal
> system in the same way that platonic realm objects are.
But I don't have a clue what is a wild-type primitive objects.
So I am not motivate in assuming that that exists (why would I assume
the existence of objects I have never see, nor comprehend in any way).
> An actual instance
> of them is axioms.
Like in the analogy:
initial condition = axioms
differential equation = inference rule
?
> 'Computation' and theorem proving from this axiom set
> is what the universe literally is...at least that is how I like to
> think
> of it.
With computation, this can make sense (through the Universal
Dovetailer).
But "axiom set" are more of the type "machine" or "theory" or
"observer" than "universe" (both in math and physics). There is no
universal dovetailing for proofs. (Only for proofs by this or that
machine).
> It's just very very messy. In local descriptions from within we
> scientists can map it into the idealised form - very useful to
> understand
> things,
After Godel we know that platonia's messiness is beyond anything (we,
or machine) can ever imagine.
> but not saying anything about what the universe is made of.
Hmmmm. The UDA shows that the "universe" is not made of things.
> The
> idealised realm helps us make conceptual sense.. but no more.
OK, but then comp is false (quite possible).
Or my reasoning is false (In that case it would be kind to give me a
clue where in the uda)
>
> I think of it like lab mice and lab flies. You make them (phenotype
> them)
> and then you get nice neat behaviour (idealised), which you then can
> use
> to contrast with the 'wild-type', which is out of your control, messy
> and
> mysterious (which is why you're doing it in the first place!).
>
> Maybe that's a poor analogy. :-)
I think so, because even just *arithmetical* platonia is full of
totally uncontrollable wild beasts. It is the fall of logicism and of
the Hilbert program. The machine which observes itself "honestly" can
discover its own unpredictibility and wildness (and free will, and
qualia, and the interfering dreams (physics), taxes, death, etc.). The
machine which observes itself enough can guess that she does not know
who she is.
Bruno
http://iridia.ulb.ac.be/~marchal/
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to everything-list.domain.name.hidden
To unsubscribe from this group, send email to everything-list-unsubscribe.domain.name.hidden
For more options, visit this group at
http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---
Received on Mon Nov 27 2006 - 10:45:07 PST