RE: Seed AI (was: How hard a Singularity?)

From: Ben Goertzel (ben@goertzel.org)
Date: Wed Jun 26 2002 - 10:07:02 MDT


Szegedy wrote:
>
> Ben Goertzel wrote:
>
> >It seems clear that von Neumann was more intelligent than Einstein by all
> >conventional standards -- yet he achieved a bit less, because of other
> >factors besides "raw intelligence".
> >
> >
> >
> You 've got an interesting measure for achievement ;) It is
> mainly viewpoint
> dependent what achievment is significant or not. E.g. the work of
> Galois was
> not considered to be signifcant 40 years long, since nobody understood it.
> The Nash equlibrium is not a high level mathematics, it is still
> considered to
> be a milestone in economics. In the communism, it would not be considered
> to be too significant. etc. etc.

Of course, there is no fully objective way to measure & compare amounts of
achievement. Einstein is more famous than von Neumann because he made a few
really crisp important discoveries, whereas von Neumann's discoveries were
more diffuse. [Godel's Theorem is an example of an Einstein-like "crisp
important discovery" in the math/CS rather than physics domain...]

The point I was making, though, is that

1) von Neumann seems based on all reports to have had more of the surface
signs of "super high intelligence" than Einstein. He was a lightning
calculator, mastered new areas of science immediately, could speak with
great precision and complexity and insight on nearly any topic, etc.

2) Einstein was in my view (and the general view) at least as "great a
thinker" as von Neumann, maybe (as is the general consensus) greater

I've seen similar phenomena in my own experience.

For instance, Jeff Pressing, a friend who recently passed away prematurely,
was just an AMAZINGLY smart guy. He spoke a dozen languages, could play 10
people simultaneously at chess while blindfolded (and quite well), could
solve really tough differential equations in his head rapidly, was a master
of several instruments and extremely good at nearly all the others (not only
orchestral instruments but West African drums of various sorts, etc. etc.).
Sometimes it scared me how smart this guy was. Yet as a scientist he was
merely *very good*. I've met other scientists who didn't seem "as
intelligent" as Jeff yet had accomplished significantly more in terms of
scientific discovery.

In our own collaborations, we worked together quite well -- although I'm
*very quick* with math & logical reasoning, he was even quicker and
considerably less error-prone; but I was always the one who came up with the
out-there, sometimes-kooky, sometimes-excellent ideas that kicked our
research projects off....

One thing this kind of observation teaches us is that the notion of
intelligence is really very subtle and multidimensional.

Of course, AI has taught us that already, in a different way -- what with
Deep Blue, Mathematica, and so forth. No doubt the advent of AGI's will
teach us all sorts of freaky things about the possible varieties of
intelligence...

-- Ben G



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:39 MDT