On 9/7/2012 3:09 AM, Bruno Marchal wrote:
On 06 Sep 2012, at 21:25, Craig Weinberg wrote:
On Thursday, September 6, 2012 2:02:02 PM UTC-4, Bruno Marchal wrote:
On 05 Sep 2012, at 17:27, Craig Weinberg wrote:
On Wednesday, September 5, 2012 10:50:02 AM UTC-4, Bruno Marchal
wrote:
On 05 Sep 2012, at 03:48, Craig Weinberg wrote:
Taking another look at Sane2004. This isn't so much as a
challenge to Bruno, just sharing my notes of why I
disagree. Not sure how far I will get this time, but here
are my objections to the first step and the stipulated
assumptions of comp. I understand that the point is to
accept the given definition of comp, and in that respect, I
have no reason to doubt that Bruno has accomplished what he
sets out to as far as making a good theory within comp, and
if he has not, I wouldn't be qualified to comment on it
anyhow. From my perspective however, this is all beside the
point, since the only point that matters is the actual
truth of what consciousness actually is, and what is it's
actual relation to physics and information. Given the
fragile and precious nature of our own survival, I think
that implications for teleportation and AI
simulation/personhood which are derived from pure theory
rather than thorough consideration of realism would be
reckless to say the least.
*Step one* talks about teleportation in terms of being
reconstructed with ambient organic materials. If comp were
true though, no organic materials or reconstructions would
be necessary. The scanning into a universal machine would
be sufficient.
That is step 6.
I haven't even gotten to step 2 yet. I'm reading "In the figure
the teleported individual is represented by a black box. Its
annihilation is
represented by a white box appearing at the left of the arrow"
from 1.
Taking this to the China Brain level, the universal machine
could be a trillion people with notebooks, pencils, paper,
and erasers, talking to each other over cell phones. This
activity would have to collectively result in the
teleported person now being conjured as if by incantation
as a consequence of...what? The writing and erasing on
paper? The calling and speaking on cell phones? Where does
the experience of the now disembodied person come in?
As you illustrate here, plausibly not on the physical means
used by the brain. Step 8 shows that indeed the physical has
nothing to do with consciousness, except as a content of
consciousness. Keeping comp here, we associate consciousness
with the logical abstract computations.
So the person's consciousness arises spontaneously through the
overall effort-ness behind the writing, erasing, and calling, or
does it gradually constellate from lesser fragments of
disconnected effort-ness?
Consciousness does not arise. It is not in space, nor in time.
Its local content, obtained by differentiation, internally can
refer to time and space, but that's particular content of an
atemporal consciousness. I would say (no need of this in UDA).
If you exclude space and time, what kind of locality do you refer to?
The computational locality used in the local universal system.
Dear Bruno,
Could you elaborate a bit more on this remark? How do you define a
" local universal system"? What is "local" for you?
In my example, a quintillion people call each other on the phone and
write down numbers that they get from each other and perform
arithmetic functions on them (which in turn may inform them on how to
process subsequent arithmetic instructions, etc). Ok. So where does
the interpretation of these trillion events per second come in? What
knows what all of the computations add up to be? At what point does
the 'local content' begin to itch and turn blue? Even if it could,
why should it do such a thing?
Because it concerns a machine looking at herself and its probable
environment.
As I think of it, a machine cannot literally "look at herself"; it
can only "look at an image of herself" and that image could be subject
to errors. This allows me to relate an image that a machine might have
of itself with the image the machine might have of another machine and
thus I found the concept of bisimulation. Additionally, we can define
cases where the image and the machine itself are identical in every
possible way and thus are one and the same thing (via principle of
identity of indiscernibles).
It seems to me that most of our disagreements flow from your
assumption that the image of a machine and the machine are strictly
isomorphic, while I assume such only in certain special cases.
Then the logic shows that it is unavoidable that the machine get non
justifiable truth.
I agree. But "what is Truth"? I see truth to be a parameter of the
degree of matching between an object and its image. It is not confined
to some single spectrum for all things. Different types of objects
require different kinds of spectra for their truth valuations and thus
are not always commensurable.
You can see that informally with thought experiment, like in UDA, or
formally in the logics of self-reference.
I have problems with the technical aspects of the UDA in almost
every step. My problem is that I do not have your set of definitions and
intuitions in my mind as I read your material. I see holes and blind
spots and tacit assumptions everywhere, but this holds true of the
material of most people that I read and so I an not upset with you for
this appearance. It could be the beam in my eye that i see as a mote in
your eye... Thus I ask you many many questions, to lern to see what you
see the way that you see it. SO far I cannot understand how it is that
you do not see the necessity of physical implementation of computations.
I am studying the work of C.S. Peirce and others int eh area of
semiotics to learn if it might help me explain the problem that I see.
Step one talks about annihilation as well, but it is not
clear what role this actually plays in the process, except
to make it seem more like teleportation and less like what
it actually would be, which is duplication. If I scan an
original document and email the scan, I have sent a
duplicate, not teleported the original.
Right. Classical teleportation = duplication + annihilation
of the original. That's step 5, precisely.
You understand the reasoning very well, but we know that the
problem for you is in the assumption.
Yes, the assumption seems to presume physicality to disprove
physicality
At some place, yes. In a reductio ad absurdum.
and presume consciousness to explain consciousness.
Yes. Like we presume (at some metalevel) anything we want to
explain (from some other realm). It is not a lott, but science
works that way. We don't know the public truth. We can only make
clear our hypothesis and reason, and propose tests.
Why not just recognize it formally and say that consciousness doesn't
need any explanation other than the experience of "this" and "that".
I am OK with this, and that is why I refer mostly to the first person
discourse than to consciousness per se.
But of course an explanation of why the 1p comes from is given.
Y4ees, we have to show how the 1p is necessary and its origin. If
it is irreducible, then a general class of consequences of its presence
is helpful for its explanation.
.
Computation seems to have nothing to do with either one of them
in comp other than the fact of the plasticity and aloofness of
comp can be seen as a sign that it is neither mind nor matter.
It still doesn't answer the question of why have appearances of
mind or matter at all?
Comp is used to formulate the problem in math. Then we can see
the general shape of the solution, which is a reduction of
physics into arithmetic, with the advantage that we get a clear
explanation of the difference of qualia and quanta. And we can
test the quanta.
I'm ok with reducing physics to math or math to physics, but neither
have any link back to experience.
Physics doesn't, except an embryonic one with thermodynamic and
Everett QM.
I have yet to see this clearly. I confess that that is cause by my
inability to read your materials that are only available in French. If
only they where in English or Spanish, I could read them... But this
does not help me with the problem that i see in the MGA and Maudlin's
argument. My claim is that Maudlin is simply wrong in his thinking about
contrafactuals. The present or non-presense of the parts of the machine
that are connected by even rusty chains makes a difference. The mere
possibility of a different configuration is something that cannot be
ignored. This is the same thing as what we see in QM, where we must sum
over all possible paths, even those that are grotesquely unphysical, to
achieve a accurate calculation of some probability distribution of events.
Computer science does, and that is what I am illustrating, notably
through the duplication experiences, and the intensional variants of G
and G*.
We are thinking of duplication differently. I wish we could revisit
the "God has no name" thread and recap the ideas we explored therein.
If there is a reason, then that reason is the nature of the
cosmos, not the filing and organizing system that indexes it's
activities.
I have problems with all three of the comp assumptions:
*yes, doctor*: This is really the sleight of hand that
props up the entire thought experiment. If you agree that
you are nothing but your brain function and that your brain
function can be replaced by the functioning of non-brain
devices, then you have already agreed that human
individuality is a universal commodity.
Why? A program or piece of information is not nothing. It
asks works, can be paid for, can be precious and rare, etc.
It can't ask for anything by itself though.
Proof.
We can't coerce data into keeping secrets. All forms of secrecy
require some kind of social control of information. Data will always
talk to strangers. (see my post today:
https://groups.google.com/forum/?fromgroups=#!topic/everything-list/L9LbbtQAN9U)
?
We are the ones to whom the significance relates.
Actually God told me yesterday that we are wrong on this. Only
the jumping spider can do that.
Jumping spiders and God are us too.
OK :)
Information is nothing but an experience that can be remembered
and transmitted to other experiencers through formation.
*Church thesis*: Views computation in isolation,
irrespective of resources, supervenience on object-formed
computing elements, etc. This is a theoretical theory of
computation, completely divorced from realism from the
start. What is it that does the computing? How and why does
data enter or exit a computation?
It is a discovery by mathematicians.
And it is a valid discovery in the context of mathematical
theory, but it doesn't translate to the realism of subjectivity
and physics.
Physics, or not physics are not among the hypothesis. More in the
questioning.
It assumes weightless computation that generates weight (for not
particular reason).
We search the reason. You say "for no particular reason" without
providing a reason.
The realism of physical weight in the universe is what I am saying is
one of the things that is not derived from pure computation.
Indeed. Such a realism is shown to be an illusion, even if a
persistent one.
There seems to be no anchoring in mass (despite info-theoretic
confusions about entropy). To comp, it makes no difference whether a
program operates on a galactic scale or microscopic scale - the code
is weightless. That is not our experience of galaxies and atoms though.
But we can justify exactly that. Those are experiences, and they
correspond to stable patterns of information, not to primitive mass or
realistic physics.
You talk, like many, like if the primitive physical universe was a
datum, and not an hypothesis. This comes from the fact that such an
hypothesis has been hardwired in our brain since a long time, for it
of course an hypothesis. Indeed, it is refuted in the comp theory. You
might think that this refute comp, but then we have already agreed on
that. That is why I say that your position is coherent; you keep
matter and abandon comp. No problem, given that my point is that *in*
comp, we have to explain matter from non material relations.
*Arithmetical Realism*: The idea that truth values are self
justifying independently of subjectivity or physics is
literally a shot in the dark. Like yes, doctor, this is
really swallowing the cow whole from the beginning and
saying that the internal consistency of arithmetic
constitutes universal supremacy without any real indication
of that. Wouldn't computers tend to be self-correcting by
virtue of the pull toward arithmetic truth within each
logic circuit? Where do errors come from?
They come from the inadequacy between belief and truth.
Incompleteness makes this unavoidable at the root, and that
is why the logic of Bp & p is different from the logic of
Bp, despite G* proves Bp -> p. G does not prove it, so
correct machine already knows that they might be incorrect
"soon enough".
Your last paragraph confirms you are still thinking of
machines and numbers in a pre-Godelian or pre-Löbian way, I
think.
I admit that I have only a wisp of understanding about modal
logic and Gödelian-Löbianian ideas, but I feel like even this
surface understanding is enough to tell me that it is ultimately
a red herring.
This is self-defeating.
Why, do you feel yourself to be defeated ;) ?
These concepts seem to just be about self-reference - maps of
maps with no territory. Great for simulating some aspects of
thought, because indeed, thinking has to do with copying copies
and intellectual grammar, but feeling doesn't.
The machine knows that, already. Feeling and first person notion
have no 3p representation at all. For logical reason, explainable
with the math above.
I agree. Why does 1p machine theory propose the existence of feeling
though?
Feeling are handled by Bp & Dt & p, and behave like feeling and
qualia. They have shape, and many attributes, but are private and
incommunicable as such, etc.
How is their "shape" quantified? Topology?
These are ways of mentioning how ideas are mentioned. In
reality, this sentence does not refer to itself. There are only
characters, or pixels, or optical phenomena here. The
significance does not arise from the same level in which it is
transmitted. This is the Chinese Room. Ceci n'est pas une pipe.
This has already been commented. You confuse the 3p
self-reference and the 1p self-reference. I think.
I don't think that I do (nor does Searle or Korzybski, Magritte...)
I don't see a reference to first person in "ceci n'est pas une pipe".
Everything is 3p, at that level. The 1p is a fixed point of personal
doubting procedures, and is not representable in any way, provably so
for machines.
But I maintain that we need to have a language that allows for the
possibility of multiple minds, even if their thought that they have
separate identities is an illusion. We cannot operate as if the mind was
a single tower.
Bruno
http://iridia.ulb.ac.be/~marchal/ <http://iridia.ulb.ac.be/%7Emarchal/>
--
Onward!
Stephen
http://webpages.charter.net/stephenk1/Outlaw/Outlaw.html
--
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.
