On 29 Mar 2014, at 10:24, Stathis Papaioannou wrote:
On 29 March 2014 19:27, Bruno Marchal <marc...@ulb.ac.be> wrote:
On 28 Mar 2014, at 23:41, Stathis Papaioannou wrote:
On 29 March 2014 03:24, Bruno Marchal <marc...@ulb.ac.be> wrote:
On 27 Mar 2014, at 18:21, Stathis Papaioannou wrote:
A functionalist could agree that a computer can replicate his
consciousness but it would not really be him. There is no explicit
or implicit position on personal identity in functionalism.
This is weird. I guess you mean your notion of functionalism, which
is too much general I think, but I was still thinking it could have
a relation with "functionalism" in the math sense, where an object
is defined by its functional relations with other objects, and the
identity *is* in the functionality.
Then "function" is always used in two very different sense,
especially in computer science, as it can be extensional function
(defined by the functionality), or its intension (the code, the
description, the "body").
Could your functionalist say yes to a doctor, which build the right
computer (to replicate his consciousness), and add enough "original
atoms" to preserve the identity? Is someone saying yes to that
doctor, but only if a priest blesses the artificial brain with holy
water a functionalist?
Can you describe an experience refuting functionalism (in your
sense)?
Just to help me to understand. Thanks.
A person could conceivably say the following: it is impossible for
a computer to be conscious because consciousness is a magical
substance that comes from God. Therefore, if you make an artificial
brain it may behave like a real brain, but it will be a zombie. God
could by a miracle grant the artificial brain consciousness, and he
could even grant it a similar consciousness to my own, so that it
will think it is me.
Hmm... OK, but usually comp is not just that a computer can be
conscious, but that it can be conscious (c= can support
consciousness) in virtue of doing computation. That is why I add
sometime "qua computatio" to remind this. If functionalism accept a
role for a magical substance, it is obviously non computationalism.
Of course, the computer or computing device must be doing the
computations; if not it is unconscious or only potentially
consciousness.
I agree. A non working computer, or a frozen brain, or a Gödel number,
cannot think (assuming comp) relatively to us (looking at the non
working computer).
The person itself might still think from her point of view, in a
parallel reality or in arithmetic, etc. But this is because in that
"parallel reality" the computer is supposed to work. If we could
freeze all instantiations of that computer, the person associated with
it would "absolutely" dead. Of course that is impossible to do.
However, it won't *really* be me, because it could only be me if we
were numerically identical, and not even God can make two distinct
things numerically identical.
Even with God. This makes the argument weird. Even if God cannot do
that. But it can make sense, with "magic" matter, many things can
make sense.
It's not so weird, since even God or magic can't do something
logically impossible like make 1 = 2,
Is 1=2 logically impossible? I doubt this. It is certainly
arithmetically impossible, but all propositions of arithmetic are
independent of most logics.
and under one theory of personal identity (which by the way I think
is completely wrong) that is what would have to happen for a person
to survive teleportation.
From her first person points of view, in some non-comp theory.
I don't accept this position, but it is the position many people
have on personal identity, and it is independent of their position
on the possibility of computer consciousness.
OK.
I think you have to specify whether comp means merely that a
computer simulation of a brain can be conscious or go the whole
way with Bruno's conclusion that there is no actual physical
computer and all possible computations are necessarily implemented
by virtue of their status as platonic objects.
It is not so much in virtue of their status as platonic object
(which seems to imply some metaphysical hypothesis), but in virtue
of being true independently of my will, or even of the notion of
universe, god, etc.
But there is the further notion of implementation. The obvious
objection is that computations might be "true" but they cannot give
rise to consciousness unless implemented on a physical computer.
Only IF you assume that one universal machine (the physical universe
or some part of it) has a special (metaphysical) status, and that it
plays a special role. Implementation in computer science is defined
purely by a relation between a universal machine/number and a
machine/number (which can be universal or not).
u implements machine x if phi_u(x,y) = phi_x(y) for all y, and that
can be defined in the theory quoted below.
A physicalist, somehow, just pick out one universal "being" and
asserts that it is more fundamental. The computationalist know
better, and know that the special physical universal machine has to
win some competition below our substitution level.
But most computationalists are probably physicalists who believe
that consciousness can only occur if an actual physical computer is
using energy and heating up in the process of implementing
computations.
Well, today we know that energy is needed only to clean the memory. A
reversible computer will need only a small amont of energy to start
the computation, but the computation itself does not require energy.
They don't believe that the abstract computation on its own is
enough. They may be wrong, but that's what they think, and they call
themselves computationalists.
Yes. You are right, most physicalists are like that. But they are
wrong. They have to believe in some magical material action of matter
which has to be non Turing emulable.
No universal machine cannot distinguish locally if they are
implemented in a brain-in-a-vat, or in arithmetic. The body by itself
does not change the computation, or it would just mean that the level
was not correct. Of course, some will just say that the primitively
concrete implementation is needed to actualize that consciousness, and
that all beings in arithmetic are zombies. But that's a reification of
a metaphysical reality, which is bad philosophy, and besides, the step
8 should explain why this still introduce some non-comp magic in the
relation between mind and matter.
Step 8 of the UDA says the physical computer is not necessary;
which is a metaphysical position if anything is.
It is metaphysical, OK, but that is part of the subject matter. But
it is not a "position" or "opinion", only a logical consequence,
with some use of Occam, to be sure, as it is a consequence in
*applied* logic. The whole meta-point is that we can do metaphysics
and theology in the hypothetico-deductive way, free of a priori
metaphysical assumption, except the "yes doctor", which is as much
metaphysical than practical.
I think the "yes doctor" applies assuming computationalism AND
assuming that personal identity is transferable, which are two
separate things.
Not with computationalism as I have defined it. Nobody would say that
he survived with a digital brain in the case his consciousness is
preserved, but not his identity. Some could say yes to the doctor,
because they want maximize the probability that their children will
keep a parent as close as the "original" can be, but that is no more
computationalism in the sense I use that term. In the literature many
definition of comp are inconsistent, and very often by implicit or
explicit physicalism.
I explicitly illustrate a case of non-comp by this one, i.e. the one
where the person keep consciousness, but is considered to be another
person. Step 5 like argument, where there is no annihilation of the
original, is often used by people to argue against comp, but still
saving MEC-BEH (Behavioral or pure 3p mechanism).
I think you could extend the fading qualia argument into a fading
identity argument. If your brain is substituted one neuron at a time,
it is hard to imagine a state of partial identity, like it is hard to
imagine a partial zombie. I think that if *my* consciousness" is
preserved, my first person identity is preserved.
However, if computationalism is true and physicalism is false then
it does also follow that personal identity is transferable, since
one can't point to a particular implementation and say this one is
me but that one isn't.
I think that this is a good point. Normally, step 8 shows the
reciprocal, it shows that computationalism (in the definition I am
using since the beginning where there is 1p identity preservation,
that is used at each step!) entails that physicalism is quasi
wrong" ( quasi wrong means wrong or invoking some reification or
ontological commitment to block a logical consequence of a simpler
theory).
As step 8 applies to the relation between comp and the plausible
reality, it cannot be 100% logical, like steps 0-7 can arguably be.
So, there are some moves making possible to say yes to the doctor and
keeping primitive matter, but such moves need to refer to invisible
inaccessible notion, non Turing emulable, nor mathematically definable
and hard to relate to the mind, especially with the "qua computatio"
of the comp survival.
Such solution are worst than invoking God to explain the universe, it
is invoking God "just" to avoid a simpler (and testable) explanation
of the (appearance of universes). It looks close to irrationalism to
me, and doubly so, with the "Everett data", where the weird also
confirms the simpler.
Bruno
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.