On 14 May 2014, at 09:36, Telmo Menezes wrote:
On Tue, May 13, 2014 at 10:29 AM, Bruno Marchal <[email protected]>
wrote:
On 12 May 2014, at 16:12, Telmo Menezes wrote:
On Sun, May 11, 2014 at 9:46 AM, Bruno Marchal <[email protected]>
wrote:
On 10 May 2014, at 12:12, Telmo Menezes wrote:
On Sat, May 10, 2014 at 8:30 AM, LizR <[email protected]> wrote:
On 10 May 2014 17:30, Stathis Papaioannou <[email protected]>
wrote:
On Saturday, May 10, 2014, LizR <[email protected]> wrote:
I guess one could start from "is physics computable?" (As Max
Tegmark discusses in his book, but I haven't yet read what his
conclusions are, if any). If physics is computable and
consciousness arises somehow in a "materialist-type way" from the
operation of the brain, then consciousness will be computable by
definition.
Is that trivially obvious to you? The anti-comp crowd claim that
even if brain behaviour is computable that does not mean that a
computer could be conscious, since it may require the actual brain
matter, and not just a simulation, to generate the consciousness.
If physics is computable, and consciousness arises from physics
with nothing extra (supernatural or whatever) then yes. Am I
missing something obvious?
Yeah, I always feel the same about this sort of argument. It seems
so trivial to disprove:
"even if brain behaviour is computable that does not mean that a
computer could be conscious, since it may require the actual brain
matter, and not just a simulation, to generate the consciousness."
1. If brain behaviour is computable and (let's say comp)
2. brain generates consciousness but
3. it requires actual brain matter to do so then
4. brain behaviour is not computable (~comp)
so comp = ~comp
I also wonder if I'm missing something, since I hear this one a lot.
I guess other might have answer this, but as it is important I am
not afraid of repetition. O lost again the connection yesterday so
apology for participating to the discussion with a shift.
What you miss is, I think, Peter Jones (1Z) argument. He is OK with
comp (say "yes" to the doctor), but only because he attributes
consciousness to a computer implemented in a primitive physical
reality. Physics might be computable, in the sense that we can
predict the physical behavior, but IF primitive matter is necessary
for consciousness, then, although a virtual emulation would do
(with different matter), an abstract or arithmetical computation
would not do, by the lack of the primitive matter.
I agree that such an argument is weak, as it does not explain what
is the role of primitive matter, except as a criteria of existence,
which seems here to have a magical role. (Then the movie-graph
argument, or Maudlin's argument, give an idea that how much a
primitive matter use here becomes magical: almost like saying that
a computation is conscious if there is primitive matter and if God
is willing to make it so. We can always reify some "mystery" to
block an application of a theory to reality.
Ok, I tried to think about this for a while. It appears that it
also connects with the issue "can there be computation without a
substrate?".
Please see what you think of my thoughts, sorry if they are a bit
rough and confusing:
In a purely mathematical sense, it seems to me that computation is
simply a mapping from one value to another.
Well, it is a special sort of mapping. There are 2^aleph_0 mapping
in general, but only aleph_0 computational mapping. So it is a bit
more than a mapping.
Any computer program p can be represented as a value under some
syntax.
Any program + some data,
Why "+ some data"? Any additional data can be made part of the
program, no?
Sure. But it helps to think in both ways.
and don't forget that the universality requirements makes obligatory
that some programs will not stop, without us knowing this in
advance. Non termination entails a lack of value, here. yet, a non
stopping programs might access computational states not met by
terminating programs.
This doesn't seem true to me. Trivially, whatever computational
state a non stopping program accesses can also be accessed by a
variation of that program that simply stops at iteration n. What am
I missing?
You are right, for the relative asteroids reason. But the FPI forces
us to consider the universal winner(s) whose measure might depend on
infinite histories. We have to reason on all computations, and the
infinite one just exists and play a role, even if the infinite
sequence of finite approximations plays the same role. That is what
make universal dovetailing possible, and it dovetails on N^omega_1^CK.
So the notion of "value" is a bit too much extensional, and miss
intensional reality (related to code and means of computation).
Because of my objections above, I'm still unconvinced of that...
Which objection? With the arithmetical FPI we are in all computations.
It makes sense that the infinite one are the winners, as the finite
one are of measure zero.
A computation is a sequence of mapping done by or in a universal
machine.
The starting one is not important, and it happens that elementary
arithmetic is already sigma_1 complete, and thus Turing universal.
But once I started from the combinators, K, S, (K, K), (K, S), ...
((S, K) (K K)), etc. Then basically we can do the all thing with some
equality rules and ((K, x) y) = x and (((S x) y) z) = ((x z) (y z)).
That's Turing universal too. You get Löbianity when you add an
induction principle on the combinators.
So, taking Lisp as an example, there is a function L such that:
L( (+ 1 1) ) = 2
By doing the computation, we in the 1p can know the value of L(p)
for a certain p. If p is:
(fact 472834723947)
Then we cannot do it in our heads. We have to have some powerful
computer, spend a lot of energy and so on.
Of course, due to the halting problem, the mapping is not
guaranteed to exist, and so on.
OK.
These mappings already exist in Platonia. Why do we have to spend
so much energy and effort to obtain some of them? This only seems
to make sense if we are embedded in the computation ourselves and,
somehow, we have to attain a position in the multiverse where the
mapping is known. Once the mapping is known, I can communicate it
to you without any further computational effort or energy spending.
I am not sure you are taking into account the FPI. "we" are embedded
in infinities of computations, and we cannot know which one. We
cannot use a god or a matter to select a reality, without betraying
the comp assumption.
But empirically, I can use my laptop to select a subset of realities
where I know the outcome of some computation. I cannot be sure that
I will be successful because I might be assassinated by ninjas while
waiting, but it seems possible to steer things a certain way.
Sure! I mean you are lucky. I don't even know if you will get this post!
But it is that lawful normality that we have to justify on some
measure on the sigma_1 closed sentences.
Of course, I bet there is a stable physical reality. But comp explains
that reality in term of measure on computations which somehow try to
satisfy you. It is a continuum.
So according to Peter Jones, consciousness is generated by the
effort of moving from one observer position to another. (even
rejecting the MWI, even in the classical world, the thing can be
seen as a tree of possible future states).
But Peter Jones point is that there must be primitive matter for
both the computation and consciousness existing "really".
The result of the computation does not change depending on when I
started it, who started it and so on.
Peter Jones would say that it does matter. There are diophantine
equation which emulate you in our galaxy, but this will count for
zero in the measure because they are immaterial and not really
existing, according to Jones.
I understand that distinction, but it still seems to me that it
increases the explanation gap without adding anything.
I agree. It is basically equivalent to "don't try to search an
explanation". But that is a normal reaction for those taken Aristotle
for granted.
This seems, as you say, as an appeal to magic. The main questions
that occur to me are: how can such an hypothesis be falsified, and
if it is true, where is the ontological difference?
By comparing the measure of computation going through my state in
arithmetic, and the measure of computations going through my states
according to the theory infered from observation.
Unfortunatelmy we cannot compute such measure, as we cannot know
which machine we are. but we can compare the logic obeyed by such
measure, and the math shows or suggests that the comp measure obeys
a quantum probability calculus, like nature seems to confirm. This
makes the quantum explained by digitalness + internalization of the
views.
If you accept comp but then make such a move, you are proposing
something that is fundamentally untestable and that leads to the
exact same consequences of its opposite.
See above.
So in fact you're saying that Peter Jones' hypothesis is already
falsified if you assume comp?
Not falsified. But naked. It is an act of introducing something
unclear and unseen (primitive matter) to freeze a conception of reality.
It feels to me a bit like the "free will" discussion which, in my
view, is solved by the simple realisation that the question does
not make sense in the first place (here I agree with John Clark).
I have no problem with free-will. With Clark's definition, it does
not exist, but with a slight weakening of that definition, it makes
sense. Not sure free-will is related to this topics, though.
I just gave it as an example of an open question that I consider non-
sensical to begin with. My view is not that we have free will or
don't have free will, it's that the question doesn't actually mean
anything.
I believe in responsibility, and in my theory of responsibility there
is a need for a theory of free will, and a simple one, even mentioned
by John Clark, is that free-will is the ability to be ignorant on
cluster on alternatives and aware we have partial local control on this.
Tell me if my short explanation did work. My hole point is that the
immaterial consequences of comp are testable, by the indirect impact
on the structure of physical reality when the laws of physics emerge
from arithmetic. Roughly speaking, we can use the arithmetical
hypostases to measure experimentally our degree of computationalism.
I believe I understand you, but "unfortunately we cannot compute
such measure" still makes me suspicious. I have a Popperian view of
science, and this still smells of an unfalsifiable claim because we
cannot compute such measures.
The theory is eminently testable, and not yet refuted, despite its
simple propositional forms.
I am Popperian in that way too. But the whole point of Platonism is
that reality is not what you wish, but the truth, which what you know
+ what you don't know. Then assuming comp, we can use mathematical
logic to measure the gap between both, it has mathematical structures.
The measure is not computable, like the knower is not definable, but
the logic of the measure one can be known, and it is a decidable
quantum like logic, and the logic of the knower can be derived, it is
S4Grz, and provide an arithmetical interpretation of an intuitionist
knower. S4Grz1, that is S4Grz but where the arithmetical
interpretation is restricted on the sigma_1 sentences, gives already a
quantization. If we follow literally the arithmetical interpretation
of Plotinus, this means that the physics of heaven is also quantum.
If we believe in the Peano Arithmetic theorems, and if we say yes to
the doctor, I think this concerns us.
Best,
Bruno
Best,
Telmo.
Bruno
Best,
Telmo.
Bruno
Telmo.
--
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.
--
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 everything-
[email protected].
Visit this group at http://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
--
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 everything-
[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.
--
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.
