On 28 Oct 2011, at 01:56, Nick Prince wrote:
The QTI, or the more general comp immortality, or arithmetical
immortality is a complex subject, if only because it depends on
you mean by "you".
Can you be more specific on this?
Well, we have discuss this a lot on this list. Once you accept the
hypothesis that we are digitally emulable, it can be shown that we
have to distinguish the first person subjective life from the
plausible third person description of the body related to that person,
and that the problem of relating those first person description and
the third person description are not yet solved. But for the
immortality question, we are obliged to consider thought experience
involving amnesia, and those experiences illustrates that the notion
of personal identity is quite relative, and, with mechanism, they
makes no absolute sense at all. They might depend on what *you* want
to consider as being *you*. You might consider to be immortal just by
succeeding to identify you with your core universal identity (the
universal machine that you are), and in that case you can consider
that you could survive a strong amnesia. Some drug can help some
people to "realize" such identification. But we are programmed by
nature to resist such identification, and to identify ourselves with
our "little ego" which contains our mundane personal histories, and
this can make you doubt that you could survive amnesia. Immortality
might be a question of personal choice. Assuming mechanism, the
question of afterlife can today be shown as being very difficult.
Indeed, mechanism breaks the usual mind-brain identity thesis, and
consciousness is related to the infinitely many arithmetical relations
defining consistent extensions of (relative) computational states. The
math leads to a sequence of open problems.
Do you know Kripke semantic? A Kripke frame is just a set (of
called worlds) with an accessibility relation among the worlds. In
modal logic they can be used to characterize modal logical systems.
The basic idea is that p is true in world alpha, if p is true in
the worlds accessible from alpha. Dually, <>p is true in alpha is p
true in at least one world accessible from alpha. For example the
p -> p will be satisfied in all reflexive frames---independently
the truth value of p. (a frame is reflexive if all the worlds in
frame access to themselves; for all alpha alpha R alpha, with R the
Sorry but I have no experience in this area but I can see that if yoU
adopt non classical logic then it opens up all sorts of
With the mechanist theory/assumption, I find it better to keep
classical logic, and to derive the non classical logic from the
intensional variants of the logic of self-reference. We have the
mathematical tools to study in a clean transparent way all those
intensional nuances (which can be proved to exist necessarily as a
consequence of the incompleteness phenomena).
It should be obvious that with the mechanist hypothesis, computer
science and mathematical logic can put much light on those questions.
But those math are not very well knows (beyond professional logicians).
Testing the consequences in reality is the tricky
part. tHE Quantum mechanical formalism has been successful in so
many respects so it gives us some confidence of being on the right
But then you do have the QM interpretation problem. The Everett theory
is based on comp (alias mechanism), and I have shown that comp
generalizes QM. A priori there are more computations than quantum
computations, but a posteriori the quanyum computations can win a
"measure battle" in the limit.
Then, as other have already mentioned, what will remain unclear
hard to compute) is the probability that you survive through some
memory backtracking. The cat might survive in the worlds where he
been lucky enough to not participate to that experience, and, for
we know, such consistent continuation might have bigger weight than
surviving through some quantum tunnel effect saving the brain's cat
from the poison. The computation here are just not tractable, if we
assume quantum mechanics, and still less, assuming only the comp
hypothesis. The only certainty, assuming comp or QM, is that "you"
cannot die. But obviously you can become amnesic of some part, if
all, your existence, or you existences. Like Otto Rossler summed up
well : consciousness is a prison. With comp, and I think with QM,
there is no escapes from being conscious, in a way or another. I
like that, but then it is a consequence of those theories.
Consciousness could be a prison yes. but MWI may be false of course,
in which case maybe not. If comp says yes it is - as you suggest,
then that's another matter. The question then is: is comp more
fundamental than QM and if this be the case,
Comp is more fundamental that QM. Yes. I argue in that direction since
a long time. It is not easy because it literally force us to backtrack
toward Plato, and to abandon Aristotle metaphysics (but not
necessarily his logic, biology, and even physics up to some obvious
correction). But the basic idea is that with comp the physical reality
is not the fundamental reality. The physical reality becomes the
"border" of the universal mind (of the universal machine(s)). To be
should there not be some
way we can utilise its predictive capabilities to distinguish (prove?)
which interpretation of QM is the right one?
Comp extends the MWI of physics into a MWI of arithmetic.
Theoretically you can derive the whole of physics from the numbers'
addition and multiplication laws by using the machine self-reference
logics. Up to now, we find a quantum logic rather similar to the one
by von Neumann and Birkhoff, but it is an open problem if it is
(ortho)modular or if we need von Neumann algebra or sub-Hilbert
spaces, finitely or infinitely dimensional, etc.
The practical weakness of such an approach to physics, is that it is
very complex mathematics.
The conceptual advantage is that we can distinguish the logic and math
of the quanta and the qualia. The qualia are characterized by
accessible but unprovable truth (for the machine) obeying to the
probabilistic variant of the provability predicate. The main variant
here is the "no-cul-de-sac-definition": that is the substitution of Bp
(beweisbar 'p') by Bp & Dt (beweisbar 'p' and consistent 'truth'). The
adjunction of Dt (diamond t, consistent 't') makes the machine
abstracting from the cul-de-sac worlds.
Of course, for the ideally correct machine, Bp implies Bp & Dt, but
the machine cannot prove that (it would prove Dt which is impossible
by the correct machine incompleteness).
I am sorry if this looks a bit technical, but in that very complex
subject, being technical is the only way to be shown wrong and to
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To post to this group, send email to email@example.com.
To unsubscribe from this group, send email to
For more options, visit this group at