From: Ben Goertzel (ben@goertzel.org)
Date: Mon Nov 21 2005 - 04:52:38 MST
> On Wed, 16 Nov 2005 "Richard Loosemore" Wrote:
>
> > would you necessarily have identified these
> > folks as utter geniuses when they were kids,
>
> No, not if the word “genius” is used as it should be, somebody who can
> make profound discoveries about the nature of reality. History is full
> of kids who preformed amazing stunts at an astoundingly early age, but
> when you take measure of their entire life they didn’t amount to diddly
> squat.
Right. But the problem is that the *potential* to make profound discoveries
is mighty difficult to measure. Someone may have this potential and then
have it not be actualized for reasons totally unrelated to cognitive
capability -- lack of motivation, unfortunate personal circumstances, etc.
It's clear that
a)
the ability to rapidly solve moderately hard logic puzzles
in isolation and under pressure
is not identical to
b)
the ability to solve extremely hard puzzles that integrate logical
reasoning with other sorts of reasoning (like information integration, and
strategic decision-making, and the creation of new sorts of concepts, etc.),
in a context of access to other humans and knowledge resources and without
a high degree of time pressure or psychological pressure
It should be obvious that the cognitive skills required for these two
things are overlapping but quite different.
The IQ test basically measures
the former whereas real-world intellectual achievement is based on
the latter. But quantifying the latter by means of a test seems far
more difficult. Any reasonable test of this sort of ability would have to
last at least a few months, and it would be very hard to control for
"cheating" of various types.
The idea of the "g factor" was in part that there is some general
intelligence
capability that underlies both a and b. The more recent "multiple
intelligences"
idea argues against there being any single g factor, and argues that b is
a result of the intersection of multiple somewhat independent cognitive
abilities,
only some of which are measured in a.
What makes things confusing however is that there is a significant
correlation
between a and b --- certainly a predicts b better than other quantities such
as
physical strength, or short-term memory size, or the ability to do rapid
arithmetic,
or grades in a typical school. (In principle, grades in school could
measure
b better than a, but that would occur only if schools were intellectually
challenging and students were motivated...)
-- Ben G
This archive was generated by hypermail 2.1.5 : Tue Feb 21 2006 - 04:23:20 MST