Le 11-juin-07, à 08:05, Tom Caylor a écrit :
>
> On Jun 10, 5:10 am, Bruno Marchal <marc....domain.name.hidden> wrote:
>> ...
>> After Godel, Lob, .... I do think that comp is the best we can hope to
>> "save" the notion of consciousness, free will, responsibility, qualia,
>> (first)-persons, and many notions like that. Tthe "only" price: the
>> notion of matter looses is fundamental character, and we have to
>> explain matter without postulating it as usual ...). We have to come
>> back (assuming comp) to Plato, or better Plotinus, Proclus, ...
>>
>
> How is assuming comp any better than believing in the personal God?
Because in general it is hard to make third person testable statements
on personal God. Also, with comp, machines HAVE TO be "theological
machine". That is, comp does not prevent some "mystical" (true but
unprovable) beliefs: on the contrary, comp makes them obligatory (at
least for the ideally correct machines).
With comp we can argue that consciousness is already such a mystical
state. It is a state such that you have "visions" making you belief in
"a reality". Even cats can believe in invisible mouse, when hunting!
The closer thing to consciousness for the lobian machine is the "state
of being consistent". With machine talking first order arithmetics, "to
be consistent" can be identified (actually by 1930 Godel Completeness
theorem) with "having a unameable reality" capable of satisfying your
set of beliefs. and "to be consistent" belongs to machines' corona [G*
minus G]. Indeed, by Godel second theorem, the machine statement "to be
consistent" is true (as we can know for simple machine) but unprovable
by the machine. After Godel we know that machine can understand/infer
that any of their beliefs in a reality has to be theological, even the
belief in a physical reality, or whatever.
Few people seems to realize the immensity of impact of Godel's
discovery (to begin by Godel himself as compared to Emil Post or Alan
Turing, ...). Before Godel, after the work of Cantor, mathematicians
were hoping to secure the many use of infinities in math by the
finistic use of their names in finistic theories. After Godel, we know
that we cannot secure the finistic realm itself and that we have to
invoke higher infinities just to talk on those finite things. Before
Godel we could have believe that the infinite can be secure by the
finite. After Godel we know we have to rely on the infinites just to
get a tiny scratch idea of what the finite things are capable of. This
has given rise to the branch of logic known as "model theory", for
example, where infinite objects are used to give clues on finite
theories.
Note that I am not equating consciousness and consistency. But I am
open to the idea that consciousness is related to unconscious
(automatic, preprogrommed) self-interrogation of self-consistency. This
makes possible to interpret Helmholtz theory of perception (as
unconscious bet) in the lobian self-referential discourses.
Because we got that "mystical state" at birth since most probably
billions years, we tend to be a little blase about it, and this
explains why we have to do some work to abstract from long-time
prejudices, but then that is what science is all about (as Plato and
Descartes have seen).
(For the "modalist", consciousness is not "Dt", but "Dt?". The
interrogation mark remind that Dt belongs to G* minus G.)
I have to go by now and I will try to explain soon why such an
inference of "Dt?" gives some advantage relatively to some very general
relative survival goal (mainly it gives a relative speed-up) ...
> Comp seems like a lot of work.
Yes indeed. Two times more work than materialist are used to think. We
have to isolate a "theory of mind" AND then, it remains to test the
physical laws forced by that theory of mind, as the UDA and the
arithmetical UDA justifies (or should justify).
But the scientific attitude always asks for "lot of works",as I just
said above.
C'mon Tom, we are not in a Holiday club here, are we?
:)
Bruno
http://iridia.ulb.ac.be/~marchal/
--~--~---------~--~----~------------~-------~--~----~
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.domain.name.hidden
To unsubscribe from this group, send email to everything-list-unsubscribe.domain.name.hidden
For more options, visit this group at
http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---
Received on Tue Jun 12 2007 - 06:35:29 PDT