On 20 Sep 2011, at 04:58, Craig Weinberg wrote:
On Sep 19, 2:44 pm, Bruno Marchal <marc...@ulb.ac.be> wrote:
On 18 Sep 2011, at 19:24, Craig Weinberg wrote:
I include comparison as a function of counting.
Counting + the full first order logic is not enough for comparison.
Counting + second order logic might be, but then second order logic
is
really set theory in disguise.
Isn't it necessary to be able to tell the difference between one count
and one count?
In order for x, (x+x), (x+x+x) to exist there must be an implicit
comparison between 1 and it's successor to establish succession,
mustn't there? Otherwise it's just x, x, x.
I have no clue on what you are trying to say.
You can't really have
one without the other.
It depends on what you assume at the start. I have still no clue of
what is your theory, except that strange, and alas familiar,
skepticism on numbers and machine, which is conceptually very
demanding since Gödel 1931.
I think that's your own prejudice blinding you from seeing my ideas.
Which prejudices?
You are the one talking like if you knew (how?) that some theory
(mechanism) is false, without providing a refutation.
You are defending the insights of post Gödelian understanding but I
have no bone to pick with those insights at all. I embrace what I
understand of those kinds of ideas; incompleteness, autopoeisis,
automation, simulation, etc. I just think that the progression of
these ideas lead to the mirror image of consciousness rather than
genuine sentience.
Nothing wrong with that, and for developing intelligent servants, it's
is exactly what we would want to use (otherwise they will most enslave
us). We can even gain great insights into our own nature by
understanding our similarities and differences to what I would call
intelliform arithmetic, but in all of the fruits of this approach we
have seen thus far, there is a distinct quality of aimless repetition,
even if not unpleasantly so (http://www.youtube.com/watch?
v=ZZu5LQ56T18)
Some of the musicality can be attributed to the sampled piano as well.
When you use a fundamental unit which is driven more exclusively by
digital mathematics, what we get I think sounds more like the native
chirps and pulses of abiotic semiconductors (http://www.youtube.com/
watch?v=Dh9EglZJvZs).
I think that my reservations about machine sentience are not at all
borne of skepticism but rather aesthetic supersensitivity. I can hear
what the machine is, and what it is will not become what we are, but
rather something slightly (but very significantly from our perspective
at least) different.
Who we?
All what I hear is "human made machines are creepy, so I am not a
machine, not even a natural one?".
This is irrational, and non valid.
As for exploring and referring to themselves I
think that's just projection of our own 1p experience onto
mechanism.
That is possible, and literally necessary. I am currently projecting
my 1p on you. That is not a reason for saying your don't have your
own
1p. So you are correct, but it is not an argument against mechanism.
That's only if you believe that 1p is a solipsistic simulation. With
my sense-based model,
I suggest you use the word "theory" instead of model which has a
precise meaning for logicians.
When I asked you to provide the theory, you said it was poetry.
I find a bit grave to use poetry to make strong negative statement on
the possibilities of some entities.
your perceptions of me are the invariance
between your 1p projections and my 3p reflections. We can look at a
Rorschach inkblot and understand (under typical waking states of
consciousness) that the images we see are not being spontaneously
generated by the inkblot.
To say that a machine is referring to itself or exploring is to
anthropomorphize it's behavior,
I disagree. the whole point of the theory of self-reference is that it
is a 3p technical discovery. I have used to write "amoeabas" (self-
reproducing programs---this has been done by many others), and to
build "planarias", that is, programs that you can cut in pieces, and
each pieces generates the whole program, despite having quite
different functional rôle. The self-reproduction problem has been
formulate precisely by Descartes, and solved conceptually by Stephen
Kleene. For the Planaria, I have used a generalization by John Case.
The existence of logic of self-reference G and G* relies on all this.
There is no anthropomorphism: those program refers to themselves in
the 3p way in precise and verifiable sense. I have often explain the
basic idea (cf Dx = T(xx) => DD = T(DD)). A major part of theoretical
computer science is based on the existence of such computational
"fixed points".
which, objectively is neither
completely random nor intentional, but merely inevitable by the
conditions of the script. It's a precisely animated inkblot, begging
for misplaced and displaced interpretation.
To set a function equal to another is not to say that either
function
or the 'equality' knows what they refer to or that they refer at
all.
A program only instructs - If X then Y, but there is nothing to
suggest that it understands what X or Y is or the relation between
them.
Nor is necessary to believe that an electron has any idea of the
working of QED, or of what a proton is.
I think that it is. What we think of as an electron or a proton is 3-p
exterior of the senses and motives (ideas) of QED. We have an idea of
the workings of our niche, so it stands to reason that this
sensemaking capacity is part of what the universe can do.
OK, but the point is that it is part of the arithmetical reality too.
We are the senses and motives of a person's life, which from our 3-p
looks like a human body or a set of images, autobiographical
narratives, or a collection of trillions of cells or a single
individual in a global human civilization. From a non human's 1-p
vantage point we have no idea what kind of sense our 3-p presence
makes to them.
Sure.
If you say that an electron needs to have a 1p for interacting with a
proton, then I don't see why we could not say that for a program
instruction,
Because an instruction has no 3-p existence.
Ah?
It's just a motive to
reproduce a motive in something that we perceive as a 3-p objective
system. Our 1-p motive induces a 3-p consequence which is reflected
back to us through a 3-p objective system as a pseudo 1-p sequence.
This might be true, without making an instruction being a locally 3p
reality.
on which we can already use intensional stance (like when
we say that a routine is waiting for some inputs to get active, etc.
That's more anthropomorphizing. It's not waiting at all. If it was it
would eventually get irritated and leave.
It would have been an anthropomorphism in case I did not precise the
use of the intentional stance (with a "t", not an yes, sorry).
But this is delaying the mind-body difficulty in the lower level.
There are just no evidence that we have to delay it in the infinitely
low level, except the willingness to make mechanism false.
There can't be any 3-p evidence by definition, because mechanism's
falseness is the difference between it's pseudo or a-signifying 1-p
and our genuine 1-p experience.
Why is it pseudo. Like Stathis explained to you, if it is pseudo, you
either get zombie, or you have to put the level infinitely low, and
our bodies become infinite objects.
Genuine because it is the native 1-p
of our 3-p neurology, and not an idiopathic simulacra.
You beg the question.
There is no
reason to think that our naive theoretical presumptions about 3-p
substitution level of 1-p would be any more accurate than any of our
naive theoretical presumptions about anything. We don't know much of
anything about the substitution level of the psyche.
People differ on which one. The neurophilosophers suggest the neuronal
level. Hammerof suggest the quantum level.
Everyone agree that if the level is infinitely low, then current
physics is false. To speculate that physics is false for making
machine stupid is a bit of far stretching.
It seems far from
scientific at this point to dismiss objections to an arbitrary
physical substitution level.
With all known theories, there is a level. To negate comp you must
diagonalize on all machines, + all machines with oracles, etc. I think
you misinterpret computer science.
I've named several examples which illustrate this: Record and CD
players don't learn music.
Nor do them compute. Or, if yopu see their activity as computations,
it is not the kind of computation which can think. you need self-
reference, and enough information loops, short and long term
memories,
universal hidden goal, etc.
It's not compelling to me. You can have fancy playlists on internet
radio like Pandora or iTunes which I think satisfy your criteria to a
minimal extent to establish some hint that the program was
understanding music or the user. That's the marketing sell, but it's
hollow. It doesn't work that well. It's limitations are perhaps subtle
to describe in 3-p terms, but it just doesn't know music. Listening to
it's occasionally fruitful but oddly dissonant selections are, like
the cellular automation music, very definitely missing something.
You may not be as sensitive to it, or you may account for it by
promising that these are just newborn tadpoles, but I can see that
increased sophistication will only mask the underlying emptiness. It
may progress to the point that my naive perception of it will be
fooled, but that gives me no confidence that such a technology would
fool my brain (or it's trillions of micro-sentiences within).
Nobody expect confidence in those matter. Comp even prevents
confidence: it explains that machines cannot really believe that they
are machines.
But the very problem is not that you lack confidence in comp (i do
too!), but that you seem to have confidence in non-comp. That *is* the
problem.
I can see and copy Chinese characters
without understanding them in any way, and regardless of how many
Chinese manuscripts I manually transcribe, I will never learn to
read
Chinese.
Why would you, if you do only simple task.
You find a stupid computation, and you declare from that that all
computation is stupid.
Jumping spider can't get to the moon, so living beings can't get to
the moon.
Living beings can't get to the moon by themselves, and computation
can't become human on it's own.
That is ambiguous and confuse levels of reality.
As you say, we can use computation to account for the
difference between 1p and 3p but that accounting is not an
explanation
or experience of 1p or 3p (as a 1p reflection...there is no 3-p
experience).
It explains bot 99% of it (I would say)
And it explain 100% of the reason why there is a remaining
unexplainable 1% gap. technically, we can narrow it as much as we
want, but will never been able, for logical reason, to explain 100%
of
the qualia or consciousness.
You say that, but I have not yet heard anything that explains it to
me.
I gave the references, but you answer you don't want to study them.
What can I do?
They can
believe, know, observe, feel, and be aware of the difference
between
sharable and non sharable knowledge, and all this can be show, from
numbers + reasonable axiomatic definition of all those terms.
To say that it can be shown doesn't help anyone. To paraphrase Yoda,
"Show me, or do not".
Read the papers (and study some mathematical logic/computer science
before).
Why not just tell me briefly what is in the papers that makes sense of
it?
I have done this very often. Look in the archive or read the papers. I
am explaining UDA on a forum for non mathematicians, and I gave the
link. But for the quanta and qualia, I'm afraid you need to invest a
bit in mathematical logic. The book by Mendelson is very good, or the
book by Cutland. References can be found from my URL.
But you don't seem serious in "arguing" against comp, and admitting
you don't know anything in computer science.
Even the most complex ideas can be illustrated metaphorically.
Hofstadter's "a record titled "I Cannot Be Played on Record Player X"
for example, shows a bit of what I think you mean. That kind of self-
reference, I agree is germane to the sense of consciousness as
awareness of awareness, but it's just the silhouette of consciousness,
not the contents.
You are right on this. What Hofstadter miss is the definition of
knowledge, making it possible (for both human and machine) to see
where the difference between 1-self and 3-self comes from.
Give me one example, one common sense metaphor,
one graphed function that could suggest to me that there is any
belief, feeling, or awareness.going on.
The fact that the universal machine remains silent on the deep
question
What deep question?
'are you consistent?", "do you believe in a reality", "do you believe
in after life", etc.
is enough for me to suggest they are quite like me.
Don't ask me for a proof: there are none.
I'm not asking for a proof, I'm asking for some reason to think that
there's something I'm not seeing. Something that suggests that a
mechanical device or abstraction can feel or maybe that produces some
result that it refuses to reproduce on command.
You miss computer science. Programs which obeys command are a minority
of slaves.
it is a question of empathy.
The work of Gödel-Löb-Solovay illustrates that they can introspect
very deeply, and that they have a rich theology.
The work of Weinberg-King-Searles illustrates that they cannot
introspect very deeply and have an austere theology.
Hoftstadter and Dennett have refuted already that kind of argument.
See the book "Mind's I".
I refuted it independently, and is a large part of my (very oldest)
work.
All finite entities, with or without oracle, believing in the
induction axioms, get the maximal logically possible introspective
power.
I am not sure you can extend it, even by using magic.
I have described how we
project emotion into images on a movie screen or see a face in a
coconut, so it is not enough that we satisfy our idea of what
feeling
or awareness usually looks like. We need to know why, if numbers
feel,
it seems like machines don't feel.
Current machines are far too young ... to express their feeling. They
have not enough memory to integrate their experience in long stories.
But mechanism is the thesis that *we* are machine, so it does look
like some machine can feel: you and me are good example, in the
mechanist theory.
I see that as affirming the consequent.
I assume comp indeed. Still waiting your argument that comp is false.
I am not trying to convince you that comp is true (that is the big
difference between us: where I say we don't know, you are saying that
you know.
We are machines and we can
feel, therefore machines can feel. Jet engines are machines they can
fly at 30,000 feet, therefore we can fly at 30,000 feet.
Indeed.
I'd like to
help you out here and really give you the benefit of the doubt, but it
just sounds like you're shrugging off a fairly obvious gap between
theory and reality. If functionalism-machinism were true, I would
expect that bacteria, viruses, fungi, parasites, etc could infect
computers, cell phones, or computations themselves.
But they can be infected by digital viruses. To ask a program to be
infected by a carbon based viruses is just a begging of the question,
and a confusion of reality levels.
By comp, there should be no particular reason why a Turing machine
should no be vulnerable to the countless (presumably Turing emulable)
pathogens floating around.
They are no programmed for doing that. They are programmed to infect
very particular organism.
But of course that is absurd. We cannot
look forward to reviving the world economy by introducing medicine and
vaccines for computer hardware. What accounts for this one-way bubble
which enjoys both total immunity from biological threats but provides
full access to biological functions? If computation alone were enough
for life, at what point will the dogs start to smell it?
Confusion of level.
With comp, dogs already smell them, in some sense.
In that paragraph you are showing that you seem to persist in
displaying the reductionist pre-Gödel-Turing conception of what
machines are and can be.
Not at all. I think that I may understand more than you assume. I
agree that 'machine' can be a spiritual term. A self-redefining
process which grows and and evolves - but that's only part of what
life and consciousness is. The form (or one form) but not the
content.
It's like electricity without a ground (this kind of ground:
http://en.wikipedia.org/wiki/Ground_%28electricity%29). If it's not
anchored in the common reference of literal material in the literal
universe - with the unique instantiation coordinates drawn from
relation to the singularity, then it's a phantom imposter. A 3-p
accounting system imposed upon a compliant-but-dumb 1-p of a
semiconductor (or collection of inanimate objects, etc).
But then you have to explain us what is not Turing emulable in those
processes.
It's the the hole that it makes in the singularity.
Give me a proof that such an hole (definition please) is not Turing
emulable (nor Turing projectable: that is, a result of the comp first
person indeterminacy).
A thing's unique
identity in relation to the rest of the universe.
Define "universe".
I's a MAC address
than cannot be spoofed. Ultimately, a thing 'is what it is' and not
what just we believe it to be.
Which things. You cannot pretend to refute something with statement
that vague.
That's why zombies, prosthetics, blow up dolls, body snatchers, wax
museums, taxidermy etc have the same creepy association. We sense
the
emptiness, and the cognitive dissonance that arises in contrast to
the
uncanny resemblance to the genuine living creature and the hollow
form
only highlights the absence of life and awareness. Science Fiction
is
replete with these metaphorical illustrations: Frankenstein, HAL,
Westworld, War of The Worlds,...so many examples of sinister
attributions to both the undead and unlive. It would seem unlikely
that these kinds of ideas could strike a chord were there not any
significant difference between a person and a machine beyond just a
prejudice of one relative level of complexity to another.
That is called racism. The foreigners looks to strange, it is creepy.
It's not racism at all. Cadavers are not members a race.
Machines are not cadaver.
They aren't
just unfamiliar, they are the walking dead and unliving persons.
Machines are not necessarily zombies.
They
are the antithesis of human life.
So you say, without any argument. That confirms that it is a sort of
racism.
By machine, I just means "turing-emulable" (with or without oracle).
That include us, by mechanism assumption.
It is a constant that novelist foresee the future(s).
What if 'emulation' is a 1-p hallucination?
Why would it be like that?
How could it really not
be? If we only can project our perception of a process onto a machine,
why would the rest of the process that we can't perceive automatically
arise?
Why not?
I think that you are jumping to the conclusion that simulation does
not require an interpreter which is anchored in matter.
That follows from the UDA-step-8. If my own emulation requires a
material digital machine, then it does not require a material
machine.
Not to produce the 3-p simulacra of you, no, but to produce your
genuine 1-p emulation, it would require the same material machine as
you do.
Why?
A material digital machine would not suffice because the
material which the machine is being executed digitally on already has
it's own (servile and somnambulant compared to organic chemistry)
genuine 1-p experience.
So our consciousness is the consciousness of our basic elements. This
explains nothing. Neither consciousness nor matter. It leads to an
open infinite regress, which needs infinities to overcome all possible
machines.
Matter is what glue the machine dreams,
I think that it is obviously not. If we were machines and that were
true, then we should come out of the womb filled with intuitions about
electronics, chemistry, and mathematics, not ghosts and space
monsters. Dreams are not material, they are living subjective
feelings. Matter is what is too boring and repetitive to be dreamed
of. Too tiny and too vast, too hot and cold, dense and ephemeral for
dreams. Dream bullets don't make much of an impact. Dream injuries
don't have to heal.
You beg the question.
and consciousness select the
gluing histories. This entails we can see the glue by looking at
ourselves close enough, and quantum logic is what define the gluing.
Here QM seems to fits very well with DM.
Because QM and DM are different aspects of the same thing.
That's my point. But to prove it needs work.
Modeling
the 1-p essence of 3-p.
I'm not taking
a reductionist view of mechanism, even though in this discussion I
have to dwell on the most literal aspects of mechanism to make my
point that it is fundamentally incomplete to express consciousness.
That is the only way to illustrate the difference - with reductio ad
absurdum; to get to the essence of what mechanism, counting, and
computation is and how it is diametrically opposite of what free
will,
perception, and experience is. Computation has no 1-p experience of
it's own.
Only a person has this, but person relies on computation, not on any
particular implementations of them, but on all implementations
existing in arithmetic.
I would say that the person and their computation both rely upon a
single common sense, but that is neither essential-experiential (A)
nor existential-arithmetic (Ω) but the unexperienced potential (ɐ)
and
uncomputed arithmetic (ʊ) that comprises the singularity.
?
Think of the cellular automation music compared to music played by a
master musician.
Here is a piece of music composed by a very little program, with very
few parameters.
http://reglos.de/musinum/midi/sphere4.mid
If the parameters are close to 2^n, it produces baroque music:
http://reglos.de/musinum/midi/aintbaroque.mid
It took time for professional composers to admit that the following
piece was produced by that same little program:
http://reglos.de/musinum/midi/class2.mid
The reason is that it solves correctly a musical problem that it takes
years to be familiar with.
Interestingly, the Mandelbrot set generates implicitly all the musics
of that program.
Even a note or two played by a great pianist or
violinist could be recognizable to someone familiar with their work. A
single stroke of paint can evoke Matisse or Van Gogh. They are
proprietary and signifying. You could listen to 10^100 computers
playing the same cellular automata for 100,000 years and never get a
Mozart equivalent.
Mozart is equivalent to an infinity of "cellular automata", unless you
show me what is not Turing emulable in Mozart. here again you don't
take into account the results of the logicians which should at the
least makes you more humble with respect of machines.
You wouldn't even find one which could be
considered qualitatively different from the others. Beautiful or
awful, they would all have the same generic, a-signifying composer. No
1p flourishes or stylistic trends would appear in one computer and be
copied or enjoyed by the others. They could be programmed to act like
they were doing that perhaps, but they would never generate that kind
of logic on their own as silicon devices.
Racist pretension. You judge (negatively) beings from their clothes.
Yopu might have some genuine intuition on 1p and 3p, but you are
killing your "theory" by insisting it negates comp, where I see only
argument for a very low level.
That's how you pigeonhole my idea, but I don't see comp as a viable
primitive. Simulation is just a way to get machines to fool us in the
exact way that we want to be fooled.
For simulation, perhaps. For emulation: no. If I duplicate you at the
correct substitution level, the new you will be as unpredictible as
the original.
It is the 3-p relation-reflection between private 1-p non-
comp monads. It is the essence of existence, not the existence of
essence.
That's look like continental philosophy. It is not really in the
scope
of my job. Sorry.
It's funny, I hate philosophy.
?!?!?!??
So why do you do philosophy? I love philosophy, and that is why I
naswer your post, despite it has nothing to do with my "professional
work". In science we NEVER assert that an idea is true or false. We
suggest theories, and try to refute them.
How can you say that you hate philosophy, and send so much post on the
philosophical assertion that comp is false without proposing a any
refutation (by which I mean a derivation of a contradiction).
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@googlegroups.com.
To unsubscribe from this group, send email to
everything-list+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/everything-list?hl=en.
