Evolution and the Frame Problem

“The world can be partitioned, described and learned about in an infinitely large number of ways.  If you let loose in the world a truly general-purpose learning device that learns entirely without constraint, it will, by chance, set off acquiring information in what is an infinitely large search space along any one of an infinitely large number of search paths.  The chances of such a device learning anything that is biologically useful within a single lifetime is vanishingly small…

There is really only way to solve…[the] almost intractable [frame] problem.  This is indeed to have vast reaches of time in which to search and the ability to conserve selectively knowledge gained over such massive search spaces.  And that is what evolution does.  It gains knowledge of the world across countless generations of organisms, it conserves it selectively relative to criteria of need, and that collective knowledge is doled out to individuals, who come into the world with innate ideas and predispositions to learn only certain things in specific ways.  Every human, every learner of any species, begins its life knowing what it has to learn and be intelligent about – we all come into the world with the search space that we have to work in quite narrowly defined.  There is empirical evidence to support this view in a whole range of species.”

Henry Plotkin, Evolution in Mind: An Introduction to Evolutionary Psychology 173 (1997).

But what if the search space was parsed by evolution, over those vast reaches of time, in a Pleistocene environment that is very different than the one we inhabit today?