Exploring Natural Learning

• I argued (first, to Papert) there was a sigificant metaphor, superior to “planning,” for understanding the processes by which a mind constructed itself; I called this “bricolage,” the term used by anthropologist Levi-Strauss for his articulation of the model in “The Savage Mind.”
• I argued (later, with Minsky) that as much as “AI” represented an important engineering and scientific goal, the only behavior one could see as manifestly intelligent at that time was human, and that it was a good idea to look closely at how children learned in their circumstances to see how they actually learned “everyday knowledge,” as the ground of common sense. I proposed “AI inspired psychology” research to extend and replace pre-computational studies of behavior and learning.
• Minsky accepted that argument, joined my doctoral thesis committee, and we agreed (as we did beforehand) that we were seeking a general theory of intelligence.
• A clever critic will notice that if intelligence depends on evolution, then it is intrinsically rooted in particular embodiments, thus unlikely to be universal in nature. (Well, then maybe this is a research issue, more than one of ideology.)

The conclusions of this study were revised and extended for inclusion in “Computer Experience and Cognitive Development.”

   I         II        III         IV  
  20         23         24         24  
 +20        +26        +26        +26  
----       ----        ----       ----  
  40         49        410         50

      9'   -->      8' 12"  
     -4"   -->     -0'  4" 
    ----         --------  
     ??    -->      8'  8"

LC1 began earlier. Both studies took place in the same laboratory, and overlapped in time. They both are, in fact, different aspects of the same effort to explore the processes underlying the development of learned, every-day knowledge.

Beginning:

An initial objective becomes active

— when its procedure appears achievable in a domain.

Outcomes:

The initial objective succeeds:

— this leads to a new objective through elaboration.

— the procedure becomes a more generalized tool.

The initial objective fails:

— this leads to a new objective through demon formation.

— the objective of the failed procedure becomes variablizied, that is, symbolically realizable; as its furthest extreme, this permits detachment from the local context and procedure which created it.

Consequences:

Any new objective can become an initial objective.

The range of a procedure’s efficacy is determined.