Other Directory Sites: SeekWonder | Directory Owners Forum
The following archive was created by hippie-mail 7.98617-22 on Fri Aug 21 1998 - 17:16:54 ADT
Julian Assange (proff@iq.org)
07 Apr 1998 16:11:18 +1000
David Honig <honig@otc.net> writes:
> At 08:16 PM 4/8/98 -0600, staym@accessdata.com wrote:
> >Perry Metzger wrote:
> >> Unless you don't believe in the Church-Turing Thesis, it is unlikely
> >> that an autistic person is doing anything a machine couldn't do.
> >
> >True, unless you believe (like Roger Penrose and many others) thet
> >consciousness is an inherently quantum effect, in which case you'd have
> >to consider a quantum turing machine.
>
> Actually, any computer that can really roll dice is not a computer.
> Turing machines can't do truly unpredictable things. Doesn't have
> to be quantum-based randomness (unless there's no other kind :-)
Consciousness (in the Penrose) sense doesn't exit, and is really a
rather an ill-posed question. Penrose's whole
digital-computers-will-never-do-what-we-can-so-suck-eggs-you-ai-freaks
argument revolves around his belief that a system which falls under
the Church-Turing Thesis would be forbidden from coming up with
Godel's famous 1931 paper and ... as a consequence humans (or at lest
Godel) must have some non Church-Turing (computational)
ability. Penrose postulates that the Church-Turing breaking power in
Godel's head came from quantum fluctuations inside neuronal
`micro-tubules'.
Even if you accept Penrose's basic argument, you have to question his
reliance Church-Turning at all. Godel as a human being was not a
closed system, nor would any computer trying to emulate him. It's
important to realise that any corpus of information which has some
computational process enacted upon it, can simply be viewed as the
saved state of some other process. Church-Turing may limit you, but
how do you know the glimmerings of the answer were not in your corpus
(state) to begin with? xoring a previously xored ascii version of
Godel's paper to produce.. Godel's original paper is certainly
possible, and I do not see why that argument can not extend all the
way from gramer correcting a draft of Godels paper to the original
draft; using as a corpus everything Godel ever read or heard. This is
an important argument; you do not quantum randomness, and you do not
need micro-tubules.
Cheers,
Julian.
The following archive was created by hippie-mail 7.98617-22 on Fri Aug 21 1998 - 17:16:54 ADT