Hi Craig,
Sorry for having take some time to comment your posts. I will be busy
the two next weeks, so be patient for possible comments.
I comment all your 3 posts addressed to me in one mail.
On 30 Jul 2011, at 15:50, Craig Weinberg wrote:
Chipping away at it.. more later.
On Jul 29, 3:51 pm, Bruno Marchal <[email protected]> wrote:
On 28 Jul 2011, at 17:41, Craig Weinberg wrote:
A machine cannot have an experience, it is the
container, it is that which is experienced,
No. The machine itself is not experienced. The experience of the
machine is experienced, and it may, or not, refer to some
representational level.
Fair enough. From a truly objective perspective though, how could the
experience of the machine NOT refer to some phenomenological level.
Of course.
To
say that it is representational is to conflate the referent and the
signifier.
Not at all. It is a bet on the invariance of our subjective experience
on a substitution level. Biology illustrates already the idea in the
language of chemistry.
Comp does not imply that everything is representational, nor that
Turing machine can simulate everything. On the contrary, some
machine's attributes are not Turing emulable.
In order for the machine to STOP, there doesn't
automatically need to exist a 1p experience of a red sign.
It depends on the machine. In the case of human, there usually are 1p
experience. I assume comp.
In our 1p,
we see a red stop sign as a qualitative image, we understand it as a
symbolic text, we interpret it as a pragmatic condition that motivates
us to respond with motor commands to our body to push the brake, the
brake stops the car.
Yes. But that is not an argument that some machine cannot do that. In
the comp theory, there is no need to eliminate the 1p experience.
Don't confuse the comp theory, and its misuse by materialists.
Through our interpretation we re-present the
signifier, which is a representation-neutral experience of presented
color, shape, size, and context. As the machine is a reverse
engineered logic, we have no reason to presume that our signifier -
the red light or sign, is presented just because a command is sent to
the processor queue to stop the car when the ccd in the camera
encounters electromagnetic changes of a particular sampled
configuration.
You are right, but this only means that we fail on the correct
substitution level.
If we are machine, we cannot know which machine we are, nor really
which computations go through, but we still face something partially
explainable.
It's going to stop the car whether there is an
experience of a sign or not. I say that there is an experience, but
it's likely not remotely like a human signifier and would compare as
one piano note compared to an entire symphony, if not the sum of
hundreds of symphonies filtered through different molecular, cellular,
physiological, neurological, and psychological audiences.
My point works even if you decided that your "generalized brain" (the
part of reality I need to emulate to get your consciousness preserved)
is given by the quantum rational Heisenberg matrix, of the whole
cluster of galaxies, at the level of strings.
but it has no capacity to
experience anything as an abstract design.
What is abstract? What is concrete?
An abstraction is an ideal teleological signifier, having no relevant
physical qualities itself but the capacity to be used as a template to
inform both physical and ideal forms. Concrete is the opposite, a
material referent which exists physically as an objective phenomena
which is subject to the teleonomy of physical, chemical, biological
consequences.
That seems quite abstract to me.
I don't buy that there is necessarily a given physical universe. It
is
only an Aristotelian rumor, based on a gross extrapolation on our
animal experience. But it fails, both on mind *and* matter.
I hear what you're saying, and I agree in the sense that from the
absolutely objective 0/∞p perspective there is no special difference
between physical and non-physical phenomena,
You miss the point. Comp shows and makes it possible to illustrate the
needs to explain how the physical arises or is build from conceptually
simpler non physical notions, already well known, which are the
mathematical relations.
but in SEE, the idea is
that existence is a relation of essential phenomena confronting it's
tail,
I think that you might confuse existence with consciousness.
I think a scientist does not commit himself ontologically, beyond the
terms of its theory.
through the involution of time-space characteristics.
This does not help.
In this
sense both mind and matter the notions of mind and matter lose all
absolute character of abstract or concrete - it is only through
perceptual relativity which the tail is assigned material qualities
from the 1p of it's 'head'. Perceptual relativity bundles the
individual piano notes of sensorimotive experience into the qualia
chords, arpeggios, and symphonies experienced by the human 'head'. The
very experience of essence seeing it's tail as not self is one of
ontological glamor. Fear and fascination at the image of meaning and
experience involuted through spacetime - decompactified as discrete,
meaningless non-experiences.
?
A silicon chip can
experience that machine, but it experiences it as a single large
molecule.
How do you know that?
I don't, but it makes sense to model it that way since we use silicon
for that reason, to tap into it's glass-like semiconductive
properties. Transparency, neutrality, reflection...the closest we can
get to a purely 'tail' material. You're right though, it could
experience anything and we wouldn't have any idea. Maybe it
experiences the ∞p omniscient perspective even, but I think
parsimony
suggests a piano note vs chords and symphonies model.
?
Maybe we make a giant cell out of a mutant jellyfish and
superimpose the machine on that - then you get a different range of
possible experiences and sensitivities.
That's unclear and ambiguous.
I'm saying we can get a better symphony out of a philharmonic
orchestra than a thousand drummers. Drums make us feel one way, a
cello feels a different way. Can you play cello with a million tiny
drums? Maybe. I don't think that the resonance would scale up the same
way. Can you make a cello player out of a trillion tiny drums? I doubt
it.
The comp problem is that arithmetic play cello, not just with a
million tiny drums, but also with a teragigamegatrillions of tiny
drums, and then 10^teragigamegatrillions of tiny drums, and that this
is only the beginning.
Yellow is not
Turing emulable and I can imagine yellow anytime I want.
That is not a reason why a machine cannot do that to.
There would have to be an explanation for why yellow should be any
more complex to create than a typical mechanical function.
You are mistaken. We don't need to understand a process to copy it, or
to isolate it with patience and time. But a computer is not an
explanation of things, it is a door to an unknown. With, or without
comp, we do have, at least, such a door in our head.
I don't see
any indication that experience of any kind can be emulated by anything
independent of something naturally capable of experiencing it.
Universal argument.
There
is no arithmetic description which could be understood by a blind
person so be able to see yellow in their mind.
You are right. But with comp what remains true is that there is an
arithmetical transformation of his/her brain so that he/she is able to
see yellow in his/her mind. The blind people does not need to
understand the arithmetical relation related to its qualia yellow
more than you need to understand the functioning of your brain to think.
Not completely abstract. A UM can't emulate a program having better
performance than it has.
Yes, it can. A UM can emulate all programs, including transfinite
hierachies of relatively more performant programs. Indeed UM are
self-
speeding entities.
Are you saying that I can build a computer out of silicon that runs a
program that runs a virtual server 10x faster than the silicon
computer is able to run?
Yes. On almost all inputs. Once you have written your code on the 10x
faster machine, I can find a better program making Babbage machine,
more rapid for all sufficiently large inputs. This is proved by
diagonalization, and can't be applied in practice, at least in a
direct way. It is Blum speed-up theorem.
I can't make a virtual server run better or
faster than the hardware node that it's running on.
You can, but for contingent reason related to your own most probable
level, you lost the speed on a finite numbers on inputs. All this is
not relevant for the reversal between physics and machine's biology/
theology.
That's unclear to me. I'm just saying that computing can't even
emulate Everything within the realm of computation, let alone in the
greater realm of sense.
I don't know where you are saying "just" that. It seems you say there
is a cosmos, and, if I am correct, you say "no" to the digitalist
doctor. It is your right. But to justify that you believe that comp is
false, you have to introduce some special non Turing emulable
components. And this looks a bit like invoking UFO to explain global
warming.
I don't think arithmetic can do anything by itself.
I think you are quite wrong on this, but we may debate on the meaning
of "doing".
When my computer begins evolving new operating systems by itself, and
when it's turned off then I would be convinced.
What make you sure that this is not possible in a few years, or in a
few billions of years. We discuss on a theoretical possibility and on
its consequence (having been clear that we would take an elimination
of person as a motive to abandon the theory).
I can appreciate a good poetical slogan to sum up a scientific
theory,
but such slogan per se cannot be taken as such as a theory.
It's not a theory, it's an observation.
I would say it is a personal interpretation of an observation.
That's a given, but sure.
That is not a given. Sometimes you talk as if you knew some truth. No
doubt that you do know some truth, but it should not be used in
arguing in those matters. The 1-p is important, but it can't do
science properly on its own. The civilized 1-p let the 3-p do the
science, but the beauty of comp relies that the 3-p can recognize the
1-p, even if only partially. The evidence are that above the Löbian
threshold, machine have a 1p and inner experiences.
There is absolutely nothing about
an animated CAD drawing of DNA which suggests it should be
associated
with anything that feels or thinks.
Movies does not think, right.
Computer does not think either. Nor brain.
People think, thanks to brain.
Right. So how does AI 'think' like a human brain without a human
brain?
AI?
The term "artificial" is artificial. And so, it is natural to use the
term "artificial", for species developing big egos.
Space is simulated, as through mutual pantomime. It's like a
decompression algorithm,
Not bad!
Thanks. I think it's pretty easy to demonstrate that space doesn't
exist, by imagining a universe with only one object in it. There can
be no change in position without some other object as a frame of
reference to create the relation of 'distance'.
Hmm...
padding the essential phenomena with null to
generate the existential experience of division. Time is similar,
only
it's like a compression algorithm, collapsing energy events by self-
significance.
That looks nice, but contains too many implicit statement making it
rather abstruse.
It's just a flagpole on the new world. I'm not the settler, or the
captain, I'm the spyglass guy.
The brain is the neurological machine through which anthropological
experience is developed. Human experience is a personal anthropology
through which the brain achieves significance.
That might be true at some level, but take for granted too many
assumptions.
Neurological machines, if they are machines at all, are Turing
emulable, unless you thing that a neuronological machine go to analog
states related to the special infinities that you have to add to make
it non-turing emulable.
What you're not seeing is that non-turing emulable is the definition
of awareness.
What makes you say something like that? There are relations, but only
relations, between the non Turing emulable, and awareness/
consciousness. But this is justified by reasoning. By making it a
definition you conflate quite distinguishable remarkable things.
It's not due to infinities, it's due to there being a
such thing as the opposite of arithmetic which cannot be represented
within arithmetic as anything. Why can't arithmetic embrace a
mathematics of it's own involution?
Give me a proof that it can't.
Where does the cosmos come from?
It has no where not to come from.
Can you doubt that it exists?
It looks like a balloon expanding to us because our bodies,
planet, solar system, galaxy is embedded on the surface, but
objectively it is not possible for the balloon to literally expand
since the singularity has no exterior. The result is, self-
involution
of emptiness. Emptiness between material phenomena is space,
emptiness
between signifying experiences (known as 'energy' in 3p exterior) is
time. This is why 'time flies when you're having fun'..
You make jumps, which are hard to follow.
That's true. Sorry. I'll try not to. These convos are getting pretty
long and I get impatient.
Therefore all we would have to do to
veryify 1p experience in something else is to make it part of
ourselves and observe the difference.
I am not sure that God can do that.
I think we could connect something to the brain which would let us do
that. Maybe even just magnetically induce it through the cranium.
But biology shows that the identity of our atoms has no role in the
building of our personal identity: metabolism changes them all the
time. The geographical-physical context changes even more.
Sure, but is there really anything that doesn't change all the time
except for our own 1p experience of being who we remember
ourselves to
be? We know that memory isn't located in specific neurons. It
doesn't
work that way.
OK. But it does not answer my question.
My answer is that we don't know enough about it, but that we are made
of materials that are made of atoms, so replacing them with an
arithmetic abstraction may not work any more than burning a log with
virtual fire.
Well, if we are machine, we never can know enough about it, but that
is why we bet.
You talk like a guru. There is only theories, and the mind-body
problem will have a formulation with respect of the theory (initial
assumption) you choose.
Sorry. I was born in California in 1968, so I may have guru in my
blood. I don't know so much about how theories work, I'm only
interested in how sense works.
That is nice, but to share with other you need theories, assumptions,
etc.
Rigor and clarity is needed, especially in the non formal context.
I don't assume infinite fine graining, I suggest ontological
incompatibility.
If you believe in a cosmos, you will need infinite graining in your
theory of mind, for relating the cosmos with your consciousness.
Why, if consciousness is not quantifiable?
Arithmetic emulates all
finite graining, and many sort of infinite one which can be shown
to
be relevant from the 1p of the machines.
I agree, but the 1p of the materials which host the machine are
never
going to be the same as our 1p unless the material can experience
the
arithmetic just like we experience it. Arithmetic itself is not an
experience. It requires an experiencer.
Where does the experiencer comes from?
The experiencer is a given.
From its 1-pov, I agree. But that does not mean that it is an
elementary reality that we need to assume in the ontology.
It's a primary vector of orientation. It
comes from the sensorimotive interior of the cosmos being twisted into
a private balloon through the time-space involution process. We see a
cell versus a molecule but the feeling of a cell is like a larger hole
through which experience can be poured compared to a molecule. It's a
metaphor - there is no hole that can be described in three dimensions,
it's a qualitative diameter correlate, like amperes, to describe the
level of experiential 'greatness' which can be experienced.
?
Second post: Craig wrote:
Diagonalization is a tool in theoretical computer science, to study
the structure of what is non computable, degrees of unsolvability,
etc. It comes from set theory, where Cantor used it to study the
degrees of infinities of sets.
Mecanical consept are immune for that tool, making the notion of *all
computation* the most solid of all epistemological realities, indeed
it makes it arithmetical, among other things.
I'm familiar enough with the Cantor set to get the gist of what you're
saying. What about non-epistemological or semi-epistemological
realities? I would define 1p as semi-epistemological.
Introducing new words only hides the lack of argument. Epistemological
or semi-epistemological will not prevent it by being used by machine,
humans, the cosmos ...
Are you saying that arithmetic is primitively real, and if so, why?
I do say that indeed.
I believe in it, and I have never seen someone sincerely disbelieving
it.
It can be proved minimal in a strong sense. You can't derive the
numbers laws from anything simpler (like logic for the early
logicists, nor physics, nor the pure real numbers, etc). So it is
certainly a good place to start.
And since Gödel & Co. We know arithmetic (arithmetical truth) is
inexhaustible, full of surprises, etc. It contains the Indra's net in
the form a fractal web where all UMs and LUMs reflect on each others.
It is full of life, and big bangs.
Are you saying that arithmetic is the only primitive reality though?
This is not the assumption of the UD reasoning. But once you get the
conclusion, you can understand that we don't need to assume more than
arithmetic, and that the existence of more is absolutely undecidable,
and cannot be used to justify any inner experience, so that the usual
OCCAM can be used to abandon the unnecessary postulates.
That makes sense at some level. You might intuit the crazy relation
between Bp and Bp & p.
The introspective machine might agree with you, if you were not so
undiplomatic as to refuse to listen to her at the start.
Haha, now who's sampling the gurujuice?
You are the one with strong assertions like "comp is false". I am just
pointing that we can very modestly already listen to what machines are
saying when introspecting. If you want really refute comp, you have to
study computer science. Up to now, it makes your argument invalid,
unless you show us the need of those special infinities, and what they
are.
To
understand addition and multiplication you have to be a person of
sufficient age and competence to do that.
You need only to be a universal machine.
Is an infant a universal machine? If so, would you say that she
understands multiplication? If so, why does she need to be taught
math?
Infant are universal machines. Teaching math, notably, actualizes
their universal feature. In fact when they get the notion of aging,
dying, anniversary, they are Löbian. Being universal they can emulate
in principle any machine, but you need to teach them many things.
Then the theoretical frame is now the problem. Isn't that what
science
is?
With comp we need nothing more than a universal system. Anyone would
do, and special one can be used to better formulate some problem, or
even for better hiding them. Caution.
Typically physicist search for the one which fits the most with
observation, but this just won't do if we assume we are machines, by
the UD Argument, notably.
The idea that comp needs nothing more than a universal system is part
of the theory of comp, right?
No. It is part of the conclusion of the UD reasoning. It is NOT part
of the comp assumption. Only a consequence, with some use of Occam.
What if that's the problem?
What problem? We want a problem. The goal is to explain that comp does
not solve the mind-body problem per se, but that it transform it into
an interesting mathematical problem: deriving the laws of physics from
a measure problem in arithmetic. Progress have been done.
The goal is to show there is a real MB problem, and that the
materialist common use of comp is just inconsistent.
That's probably exactly what classical physicists said at the time.
I'm the one exploding assumptions in this case though. I'm saying
that
we don't have to look at the universe as a 3p phenomenon
pretending to
be 1p - it's literally both.
That is like equating two mysteries, to pretend there is no one.
The appearance of mystery comes from our lack of an impartial vantage
point. Our 3p is not objective, it's 1p intersubjective.
The physical apparent 3-p is indeed a real first person plural view.
Again, the machine can justify what you say (accepting the classical
theory of knowledge of Plato).
Perspective would not be imaginable were it not experienced first.
Why? This would lead to infinite regress if that were true.
Your view takes the relation between 1p and 3p for
granted..
What? The study of that relation is made extremely non trivial in
comp.
Both with and without the classical knowledge theory.
Maybe I don't understand enough about how comp us used to understand
this.
I do think you should study comp. Note that this list is in advance
compared to the literature: most computationalist philosophers still
believe that comp can assume a primitively material universe (having a
role in consciousness). That simply cannot work. UDA refutes this. And
AUDA shows how and why, despite no basic universe, most UMs and LUMs
will develop stable believe in stable local realities.
Why would it be a reduction. Physics does not go very far in the
mathematical complexity. I think you confuse truth and the theories
which attempt to put some light on truth. In math, those things are
very deeply different.
They are reduction because the experience of listening to music is
significant without a formal mathematical analysis of it's structure,
while the formal analysis is not significant except in it's relation
to the 1p experience of the music.
You beg the question, you just assume that the formal coupling brain-
music is not enough. You need some infinite phlogiston in the brain to
assume this.
Third post:
So you assume that there is a cosmos, that there is inner
experiences,
and some quasi panpsychic link between them. That is also like
assuming a solution of the mind-body problem at the start.
I don't see how this could explain what is cosmos and matter, what is
mind, and what is the nature of the relations between them.
I wouldn't say there is 'a' comsos, I would say there is comsos -
order, experience.
I am a simple mind. If you assume A & B & C, it seems to me that you
do assume A.
It does explain the relation between matter and
experience,
You say so. Where is the explanation? What we already know is that you
have to speculate some special infinities. So we have not see an
explanation, but only a complexification of the problem, without
motivation (except preventing computer to have inner experiences).
one side is the opposite of the other in a continuum of
sense.
?
Sense is the relation of the two sides of the continuum.
That seems like the billionth rewording of a version of the identity
thesis.
The
nature of that relation depends entirely upon the scale and
complexity, history, purpose, context etc.
Up to infinity.
I don't think neurology has put any light on the "interior of our
minds". Only on its possible low level implementation.
Through neurology we understand the effects of neurortransmitters,
hormones, etc. How addiction works whether it's gambling or cocaine...
Lots of things.
Using comp, indeed. But that is the "easy problem". It does not touch
the problem of the inner experience. Indeed it is a software problem.
A priori, neurons have no big role, per se. Understanding how Deep
Blue works does not need the understanding of transistors.
Sure. By definition a bison is three dimensional. But a two
dimensional computer (that exists) can emulate a three dimensional
computer, in which bison can evolve and eat grass.
That's true, but it still truncates the interior dimensions of the
Bison's experience.
How do you know? You can be sure on this only if you are sure that
some special infinities have a role, but you have not show them, nor
explain their role.
Cells are made of molecules. The feeling of
cells are made of the feeling or proto-sensorimotive events of
molecules.
For which you need to speculate on a new physics, as we have already
agree.
If you look at just the unfeeling side, then you see only
physical matter, but the physical matter that you actually are
undeniably feels.
I deny this. Not only the matter my body is made from does not feel,
but, I think it does not exist basically.
Note that some experience show that we can imbue feeling on arbitrary
object (like a plastic) hand when manipulated in some way.
You stub your toe, and you feel pain in your toe.
Because I am well programmed, but you can make me feel I stub my toe,
by activating directly electrodes planted in my brain.
It's your toe. It's part of you. It feels.
Have you read about phantom limbs? Some people can have toe-ache
without toes. would you say that "nothing" feels, in this case. We
need sensorimotive insistence of the void.
Our consciousness is quite fragile compared to the robust
physical systems around us and we see that small changes in
functionality of the brain have tremendous effects upon human well
being.
Small change in any machine can make them crash.
Then it would follow that they would be careful not to crash. You
can't be careful if you can't care.
Sure, but I don't see the point, unless you beg the question again,
and assumes that a machine cannot care.
Remember, consciousness is subtractive. When the body dies,
consciousness isn't lost, it just has a new view as a consequence of
losing it's material filter. Sense is about pulling wholes through
holes. Without any resistance, there is only the whole and no need for
pulling.
I can agree, but how can you be so sure that the material filter is
not Turing emulable. It does look like a form of racism.
The word "function" is very tricky. Either you see a fnction
extensionally as a set of input-output (behavior), or you see a
function intensionally (note the "s"; it is not a spelling mistake!)
as a set of recipe to compute or to process some activity (leading to
output of not). Comp needs the two notions, and the second one is
used
(sometimes implicitely) in the notion of substitution level. A copy
of
a brain is supposed to preserve a local process, not a logical
function.
Like what kind of local process? Membrane transport? Action
potentials?
You decide with your doctor. I don't care about that. Comp says that
there is some level, not that the level is this or that. Comp explains
we can never know for sure what is the level, except indirectly by
deriving physics from what we can see when we observe ourself below
that level.
Machine do make that difference already. This is explained in full
details in some papers. Of course some amount of computer science
needs to be studied.
What would be a common-sense real world example of this?
A human.
Yes, that is the goal. Understanding what is the cosmos, where it
does
come from, what does it hurt, etc.
Anywhere that the cosmos could come from would also be the cosmos,
wouldn't it?
It is not. It would be misleading to call arithmetic a cosmos, which
is a term I prefer to use to denote the physical local reality, and
which emerges, in a very precise (and thus testable) way, from a tiny
part of arithmetic.
You are confusing a menu, with a program running relatively to some
universal numbers.
A program running relatively to some universal numbers is still not a
meal.
But this is what you fail to explain to us. How could a universal
number distinguish between a simulated meal and a meal? To be able to
do that, you need to introduce infinities and non Turing emulability
in the working of the brain.
This is like complexifying the data (in which such infinities does not
occur) to avoid a theory. You make a problem just more complex to
avoid a possible solution.
The point is that it does not fail on that. Once you accept the idea
that you are a finite machine, arithmetic explains where the coupling
consciousness/realities come from. It even explains why we feel that
there is an insurmountable gap, and why that gap has a very special
shape and its role in the apparition of the cosmos.
You can model it that way,
It is not a model. It is the consequence of a simple model (comp).
and it succeeds in it's own terms, but it's
terms are distorted
What?
to fit the ontology of it's methods.
It is the contrary. The ontology is simplified, to fit the theory.
That happens all the time in the scientifif endeavor. It is seen as a
progress, because we need to postulate less axioms, to get more
explanations.
It's not a
matter of being finite, it's that we have finite aspects and
transfinite aspects.
Like all machines.
It sounds like a good model, but ultimately it's inside out. Pain
cannot be generated within a mathematical theory. Arithmetic cannot
care.
And woman cannot reason ... We heard so much ridiculous dismiss like
that. Why not being agnostic on those issue. Build a clear non-comp
theory, and show us something which cannot be explained by computer
science, and then we will be interested. But if you say "a machine
cannot do that", it will just look the terribly sad feeling of
superiority that some humans can have from time to time.
Yes machines, in the 3p sense, cannot have experience. Nor can a
human
brain, or anything conceived in a 3p view. But machine have natural
1p
views on themselves, and it is the 1p which is subject to experience.
I agree. Our only difference I think is that you are saying that a 1p
experience of a simulation is going to be the same as the 1p
experience of the original if you copy the 3p view.
At some right level. Yes.
I say plainly not,
as a mathematical model of fire does not produce heat, not even within
the 1p experience of the simulation world.
How do you know that?
The experience of heat is
different than the experience increased velocity and collisions of
virtual atoms.
Sure, but the effect of virtual velocity and collision of virtual
atoms can have the correct effect on the virtual neurons of simulated
people, so that they will behave like if they were suffering from the
heat. If not, it means that you put something infinite and non Turing
emulable in the human brain, and this makes both matter and mind, and
their relations, artificially more complex.
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 [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/everything-list?hl=en.
