Re: Is universal artificial neural network possible?

On Sat, 10 May 2008 11:28:22 GMT, Kenneth P. Turvey wrote:

On Thu, 08 May 2008 10:10:57 +0000, Dmitry A. Kazakov wrote:

On Tue, 06 May 2008 03:52:01 GMT, Kenneth P. Turvey wrote:

On Tue, 29 Apr 2008 10:40:53 +0000, Dmitry A. Kazakov wrote:

Second, since we know how long it has taken humans to evolve, we can
develop at least a general idea of the likely complexity of the
problem.

No, because the medium is different. For example, it is known,
statistically, how long it would take for silicon atoms to form a
transistor through thermodynamic movement of particles. That does not
estimate the complexity of producing Pentium IV chips.

It certainly puts an upper limit on it, which is what I was saying.

1. To know an upper limit of complexity is useless, unless it is not proved
the least upper bound. For Pentium it is not, what makes you think it is
for "intelligent"?

2. It is incomparable anyway, because, as I said before, evolution is not
engineering. "Complexity of evolution" has no meaning for engineering and
reverse.

How so? Evolution is a statistical process. We have just one sample of
evolutionary process leading to us. Statistically, this shows perfectly
nothing.

So you can learn nothing from statistics?

Nothing from *this* statistics.

It certainly puts some time constraint, statistically. However that is
probably irrelevant as the event already happened - we call themselves
intelligent. A real constraint for Turing-complete systems could exist
if our brain used some incomputable elements.

Ok, to me this is silly. Just my opinion. If it was not computable,
then our brain clearly couldn't compute it.

This is a different thing. Real-time clock is incomputable, yet any
modern CPU has it. The very idea of an incomputable element is to
compute something, which cannot be computed otherwise.

This just doesn't follow your assertion above.

The brain cannot compute 10341st root of pi, but it certainly can read a
result of such computing. When comparing computational power required for
intelligence we should clearly distinguish the system itself and the tools
it uses, even if these tools are products of its functioning. You certainly
would not allow driving cars in athletics...

Maybe we could state it as an optimization problem if we knew more
about what intelligence is, but we didn't so far. We also know nothing
about the complexity of the problem if stated in this form. Evolution
is solving a completely different problem and the best solutions found
(bacteria, insects etc) aren't any intelligent.

We can state it as an optimization problem already. We do it all the
time. That's what standardized testing is all about. We do it in many
other ways too. When you got your drivers license you passed a small
portion of a human intelligence test. We could very easily put
together 100 of these and come up with a metric for what it means to
have roughly human level intelligence. Not only would it not be hard,
most of the work would already have been done.

That is not enough. You have to show existence of an optimum, or at
least a countable number of local optimums.

We know that human level intelligence exists.

What is the set, where the solution is looked for?

We may not have a good scientific definition of what intelligence is,
but we have a very good working definition of what it is.

It is possible that a scientific definition does work (impracticable),
but an unscientific one certainly doesn't.

So the fact that I can give you a good metric isn't sufficient?

In which sense is it good? Let a value is 345. What does this mean? It
isn't even well-ordered. Apart from that 95% of people would claim their
dogs more intelligent than the boss.

You must
also have a metric that is objective and contains no noise? That would
be quite a surprise to many in the evolutionary computation field.

No, I would like to know the properties of the metric and of the noise.
Whether something like the law of large numbers is true, etc.

--
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de

[ comp.ai is moderated ... your article may take a while to appear. ]
.

Relevant Pages