On 06 May 2015, at 14:28, Bruce Kellett wrote:
Bruno Marchal wrote:
On 06 May 2015, at 04:19, Bruce Kellett wrote:
Counterfactual correctness has not been shown to be necessary --
it is just an ad hoc move to save the argument.
Counterfactual correctness is the bone of what *is* a computation.
To have a computation, you need a universal system capable of
understanding instruction of the type IF A THEN B, ELSE C.
The local truth of the C act must be "caused" by the local falsity
of the A predicate. The computation is in the semantic of those
type of truth, at some level description of yourself.
This is not necessary for computation.
I am afraid it is, at least with the standard notion of computation. I
am not sure what you mean by computation, and from other post I doubt
it is related to the standard notion.
It would occur only in a program that required branching at some
point if the input at that stage differed. Computation is perfectly
possible without this requirement. If you have a simple linear
program that computes an output for each input, then a recording of
the action for any particular input, when replayed, would
reconstruct that computation exactly.
I disagree with this, but I will not insist as it is probably not
relevant for the argument, given that MGA applies to computation which
*might* be slightly more complex than the one described here.
But I disagree because to have a computation, you need a universal
number implementing it.
With the movie, there is no universal number needed, at least to run
the initial computation (Liz made that point too). If the movie is a
computation, it is a "linear simple program" (trying to be open to
your saying), which does not compute the kind of relation to have
consciousness.
Counterfactual correctness is not required in such simple cases.
So, such simple case are not conscious, in the best attempt to guve
sense to your notion of computation (which is not the standard).
And likewise, it is not required in more complicated situations,
such as where there is a loop, say, that requires different actions
on different iterations of the loop. The whole calculation, and
hence its recording, follows all these iterations, and the recording
reproduces them all exactly.
It reproduces exactly the physical activity which is contingently
related to the physical implementation of that computation. But the
MGA and maudlin just show that such a physical activity can be changed
quasi-arbitrarily without changing the computation, just by changing
the universal machine running it.
This that consciousness, like we can intuit, is not even in the
abstract computation, but in its semantic, which in this case refer to
alternate branching (realized or not).
If this program instantiates a conscious moment, or a whole
conscious life, replaying the recording recreates that moment or life.
In real time? Imagine the Nazis would have filmed the neural-glial
processing of the brain of their victim. Are you saying that playing
the movie at some time in some place would enacted the suffering at
place and time, and so that such movie should be illegal and
considered as torture?
Just as a recording of an orchestral symphony reproduces each bar of
the symphony as well as the whole, following exactly the fact that
each instrument plays different notes and sequences of notes in
different contexts in the score. Conterfactual correctness is just a
distraction.
You need it to have a semantics for the "if then else".
You need a computer.
But you can fix one universal system, and define all computations in
that one.
I might come back on the combinators, as they are simpler to
illustrate some points, here.
Without the notion of counterfactual correctness (or incorrectness)
there is no remorse, no regret, no guilt, no will, no hesitation, no
planning, no learning, ... You might be on the slope of consciousness
and conscience elimination.
Are you OK that the consciousness phenomenon is real and that you are
conscious, and not a p-zombie?
Are you OK that consciousness is undoubtable, yet unprovable, even
undefinable?
Well, you do believe in consciousness as you agree with "comp1", like
John Clark, and others.
Comp1 involves the notion of consciousness, by assuming it invariant
for some digital substitution.
You can't question the actors in a James Bond movie and expect to
get anything sensible, of course. But then, no one is suggesting
that a movie of someone's face records the basis of their
consciousness. The movie in question is a recording of the basic
brain processes (at the necessary substitution level). This, when
replayed, recreates the conscious moment -- not a new conscious
moment, as you point out, but a conscious moment nonetheless.
The existence of the movie (perhaps with the checking that it *is*
a computation) might be used to prove that the computation exist,
and consciousness can be associated ... with the computation, but
not with the description of the computation itself. But physical
syupervenience would imply that, and so it is just wrong that
consciousness supervene on a brain or a computer. It supervenes of
the mathematical computation that a physical computer can
incarnate, if the physical is the winner on the sum of all
computations below the substitution level.
If it did not, then the original comp argument fails -- we could
not replace all or part of our brain with a device performing the
same operations.
We did assume that a computer is needed, for the local
manifestation of my consciousness. At this stage, assuming it is
has to be primitively physical is begging the question.
What does beg the question is your assumption that the physical
substrate, be it primitive or not, can be dispensed with.
<sigh>
There are two things.
1) the mathematical facts, well known by the experts (who even asked
me to suppress any explanation on that as it is trivial for anybody
having grasp the ten first hours of course in that matter) that the
notion of computability is mathematical, with CT, and even
arithmetical, actually sigma_1 arithmetical. here we dispense with
physics, like we dispense with physics in any branch of math, except
of course when they are used or applied to physical problems.
2) Neither comp, nor MGA ever *assumes* that the physical substrate
can be dispensed with.
Comp, comp1 if you want, entails, not that we can dispense of physical
substrate, but that we cannot use it to select consciousness, without
adding magical actual non Turing emulable abilities (nor FPI
recoverable).
You remind me that I have to comment a post by Liz, which compared
Napoleon/Laplace with this, but she was not correct. Laplace said to
Napoleon that he did not need the hypothesis of God. Comp says that it
cannot uses the hypothesis of Matter, without making it a God-of-the-
gap. Laplace said ~[] (God is not necessary). Comp says here ~<>
(Matter is not possible, defining matter by what is observable).
Comp makes primitive matter into a phlogiston, like special relativity
makes ether into a phlogiston.
Assuming comp, or more exactly comp? (you need to be aware of the
hypothetical character, comp might imply that the "believer " in comp
go to hell!)).
MGA does not establish that either the original computer, or the
recording of its operation run on another physical device, can
simply be disregarded.
I think MGA does it once you understand that for a mathematician the
notion of computability, based on CT, is a mathematical notion, and
actually an arithmetical notion.
You are the one invoking some God which has to intervene to make a
computation more real than others.
And so avoid the real problem confronted by people taking seriously
enough comp (comp1) (CT+YD): we must get an explantion why some
computations seems more probable than aberrant dreams or white noise.
And that problem is a problem in math, and I give a way to handle it
which preserve the diffrence between the justifiable and the no
justifiable. But you overlook that.
If you take away both the physical device and the record player,
then you no longer have the conscious moment and/or life.
Well, then there is a flaw in the argument. But you don't seem to
study the argument.
Minimally, you can see UDA as a failed attempt to refute comp. And
AUDA as a refutation that UDA refutes computationalism. But AUDA is
constructive and gives the logic(s) of the observable (assuming only
RA and some general definitions).
Is it not nice (at the least) that the qubit to bit path is a two way
road?
Bruno
Bruce
--
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 everything-list+unsubscr...@googlegroups.com.
To post to this group, send email to everything-list@googlegroups.com.
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 everything-list+unsubscr...@googlegroups.com.
To post to this group, send email to everything-list@googlegroups.com.
Visit this group at http://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.