Intelligence

[Robert Jasiek <jasiek@xxxxxxxxxxxxxxx>]

A human's intelligence 
     is the ability to apply knowledge.
A computer's artificial intelligence 
     is the ability to apply knowledge.

Although the definitions are slightly simplified, it is illustrated
that there is little difference between humans and computers. What
does this tell us? We need to understand the following:

1) What knowledge may exist?
2) Which types of knowledge do exist?
3) How is knowledge learned?
4) How is learned knowledge applied?

1
This is go theory.

2
- rules (ex:a group is alive IF it has 2 eyes)
- patterns (ex:ponnuki)
- structures (ex:prospective territory)
- values (ex:territorial points)
- combinations of the above (ex:IF a <pattern> is a ponnuki, it has
the <structure> prospective territory and the <value> 1 territorial
point)
- others?

3
- hardcoding
- neural net teaching
- sample databases
- experience databases
- others?

4
Use search and (3) to formulate (2) as a description of (1).

***

Why am I telling you these trivialities? IMO, we should get rid of
the myth that intelligence includes feeling and intuition. Rather
it is just knowledge representation and application. The word 
intuition merely hides the fact that 10^5 or 10^6 rules are 
necessary for successful application. The art of artificial
intelligence then starts with structuring such rules -
knowledge about knowledge.

What do I do as a human when applying "intuition"? Take a hare
nakade, e.g. I immediately have the intuition to occupy the vital
point. However, this intuition is only based on knowledge!

***

Divine intelligence
     is the ability to unify knowledge representation and application.

Example: Select the right function f. Then f(situation) = perfect_move.

-- 
robert jasiek