Re: What should I say to the post-grads about Godel?

On Jun 17, 2:16 am, Marshall <marshall.spi...@xxxxxxxxx> wrote:
On Jun 16, 2:29 am, Rupert <rupertmccal...@xxxxxxxxx> wrote:

On Jun 16, 1:43 am, Marshall <marshall.spi...@xxxxxxxxx> wrote:

What to say of Penrose? He is by all accounts quite a physicist, and
yet, does he have any background that should lend any weight to
his arguments about the nature of the mind? I am not aware of any.
Also, we have the fact that his motivation is more easily attributed
to personal aesthetic feelings than by any deep crossover from
the realm of physics. Finally, we may consider that his conclusions
are ridiculous.
[...]
"Ridiculous" seems like a rather harsh word. I don't think that the
picture that Penrose suggests is ridiculous, just inadequately
motivated.

I guess you're right. I suppose I myself have some emotional
attachment to this question. I'd really *like* it to be the case
that humans are magic and special in the universe, and have
an immortal soul and all; it's just that the evidence is
overwhelmingly against that idea.


That wasn't really where Penrose was coming from. You might want to
take a look at his book "Shadows of the Mind". There is a good set of
articles about it here:

http://psyche.cs.monash.edu.au/symposia/penrose/

Basically, in Chapter 1 Penrose gives his own summary of the debate
about artificial intelligence and the various positions available on
the nature of consciousness. Then in Chapter 2 he gives an exposition
of a version of Gödel's theorem which is also quite closely related to
Turing's result about the non-solvability of the halting problem.
(This is a book which attempts to reach a large audience). He then
tries to show that from this theorem we can conclude that "Human
mathematicians are not using a knowably sound algorithm in order to
ascertain mathematical truth". I myself would think he's got a pretty
good case here. He then spends the rest of the chapter addressing
various objections. Then in Chapter 3 he introduces further arguments
which attempt to show that human mathematicians are not using *any*
kind of algorithm in order to ascertain mathematical truth. Some of
the arguments are quite novel. This chapter ends with a fantasy
dialogue beween a hypothetical intelligent mathematics-performing
robot in the future and the robot's creator. The robot's creator
presents the robot with Penrose's arguments and the robot goes mad.
I'm inclined to think the arguments of Chapter 3 don't hold water, but
as George Boolos once remarked, at least it is a point in Penrose's
favour that people have a hard time agreeing exactly what is wrong
with them. We talked the matter over on sci.logic a few years ago and
Daryl McCullough had some interesting perspectives which I thought
were quite useful. I planned to write something for publication on the
topic myself but I haven't got around to it yet because I've been busy
working on a Ph.D. thesis. Anyway, that was as far as I got in the
book, but after that, once Penrose thinks he has shown that the human
mind is non-algorithmic in nature, he then goes on to talk about
physics, and the philosophy of quantum theory, and microtubules in the
cytoskeletons of neurons, in order to outline a bold conjecture for
how future developments in the theory of quantum gravity will explain
how the human brain is able to act in a non-algorithmic way. I didn't
read this part of the book. Solomon Feferman said it struck him as
"quixotic".

So Penrose does think that consciousness will ultimately be understood
by the methods of mathematical physics, but that it will turn out to
be non-algorithmic in nature. You could say that in a sense this is
all just another manifestation of our desire to believe that we are
very special but it's a little bit different to that old idea of an
immortal soul. It's a 21st-century attempt to put humans on a special
plane.


If the mind is not the result of computation, what is it the result
of?
There can be only two types of answer: 1) is to say that it is some
physical process that we are not yet aware of, and 2) is to say that
it is supernatural in nature. 2) is not worth refuting; we might as
well just say consciousness comes from Santa Claus.

You can say that the mind is some sort of physical process which can
be understood by the methods of mathematical physics, and this is what
Penrose believes (though he adds that the process is not computational
in nature).

Saying both of those things at once strikes me as contradictory. Or at
least a contradiction of Church-Turing. And saying the brain is not
computational is just emptily perverse. What aspects of the mind
are not information processing? There are none.


Well, I guess you need to read Part II of "Shadows of the Mind" and
his speculations about quantum gravity and the microtubules in the
cytoskeletons of neurons.

You could also say that the mind is some sort of non-
physical entity (you would not necessarily have to speak of anything
"supernatural", and by the way I have to confess that I don't know
what the word "supernatural" means).

Certainly the *brain* is a physical artifact. Is there any question
of whether the mind is a manifestation of the brain? The mind
is non-physical without being supernatural.


Penrose defines four positions about the nature of consciousness, A,
B, C, and D. If I remember rightly D is the position that
consciousness cannot be understood by the methods of mathematical
physics at all, and C is the viewpoint of strong AI. He rejects both
of those.

You can't just say that's "not
worth refuting" and like believing in Santa Claus. It's a live
philosophical debate and there are impressive contributors on both
sides

I dunno. Are you a priori dismissing my use of a priori dismissal?
I acknowledge Penrose for example as a impressive intellect,
and yet in the field of AI he just looks like Bishop Wilberforce
to me.


I could understand someone thinking they won't take the trouble to
study Penrose's thoughts about AI in too much depth, but to say that
any position that maintains that consciousness is not fully reducible
to physical processes is "not worth refuting" does not seem correct to
me. The hypothesis that the universe can be understood in terms of a
purely physical ontology (which I think Penrose agrees with, by the
way) is a live research programme which has had a lot of successes,
but that doesn't mean that any position which says we should try to
construct an alternative research programme is not worth refuting.

[snip...]


That is because I myself have not read what they have to say. I have
read a sketch of what Unger says about the matter and it seemed to me
to be problematic because he seemed to be placing too much weight on
the hypothesis that we have free will.

Here I run in to a problem, because I don't know what "free will"
means.

Unger has trouble working out what it means, too. Because the
alternatives appear to me that either determinism is the truth, or
else determinism combined with elements of pure chance is the truth,
and on either conception there seems to be no room for anything we
would want to call "free will". He tries to make a case that there is
a third alternative. I'm sure you could be forgiven for feeling
unclear about what he's talking about.
.

Relevant Pages