[ExI] intelligence and generalization

[Stuart LaForge]

Bill Wallace wrote:


> Stuart, I hope that creativity is not limited to the arts.  Why shouldn't
>  an AI create math?

You got me there, Bill. Math, or pretty much any discipline really, can be
raised to an art form by those who have mastered it. I could argue that
mathematical realists, as opposed to mathematical constructivists, contend
that that mathematical theorems are objectively real and are therefore
discovered rather than created.

But I seem to recalling reading that Rodin, or one of the other great
sculptors, said, and I am paraphrasing here, that the secret to their
success was that at the outset they could see the statues trapped within
the block of marble and they simply "released" the shapes already within
so that everyone else could see it too.

> 'Physiologizing' is what we in psych call reducing everything to actions
> in the body/brain.  So you know that a certain part of the amygdala
> controls aggression.  So what?  Destroy it, stimulate it, and you get
> predictable results.  But how does that relate to what is happening in a
> complete body and brain?  The only answer I can think of that has any use
> at all might be to develop drugs that act on the brain part.  Do we really
> want our lives to be determined by a host of drugs we take in to activate
> or inhibit brain functions?

A significant portion of people already take a host of drugs, either
self-medicated or prescribed for recreation or some chronic condition or
another. How is a drug that affects the brain any worse than a shot of
whiskey or a drug that controls blood pressure?

> In extreme cases, yes, that could be
> appropriate, such as in a person who goes into rages and loses control.
> That person could face a
> decision:  go on the drug or go to jail.  But in ordinary life?  Not for
> me.  Not unless functions are lost or become limited - Cialis comes to
> mind!  (I am turning 77).

Societal norms have shifted throughout history. Sometimes a past
generation's extremes become a future generation's norm. If population
densities start to approach one person per square meter, we might need to
go on some equivalent to Huxley's Soma to keep from killing each other.

If you still find erections useful at 77 then my hat's off to you, sir. :-)

> As for consciousness, we can find the brain parts that control it and
> thus control those brain parts.  But as for defining and copying what
> those parts do - I have serious doubts.

The complexity of consciousness is such a beast that we might have better
luck just copying those parts without trying to define their function.
Also consciousness might be a function of both body and environment rather
than simply body. Maybe consciousness is less self-contained in the
brain/body and more dependent on environmental inputs than we think it is.


Stuart LaForge