This may be a little out of context, but it was the first thing that popped
into my mind when I read this.
In the last year or so, a couple pieces of pretty sophisticated image
processing software were produced that could generate fairly detailed 3d
objects and spaces by analyzing normal 2d photographs. Some of this
software is quite good at making educated "guesses" as to the nature of
parts of objects that are not explicitly shown in the photographs. Of
course, the more different photos you have, the better the model that is
constructed. All in all, the best of this class of software produces very
impressive results.
The point of all this is that I see a clear convergence between the ability
of computers (given the proper software) and the ability of the human mind.
The description of human visual processing quoted above, is noticeably
similar to the computer image processing software I mentioned below it.
Given the abundance of detailed 2d imagery of the earth's surface and
man-made structures and objects, it would be possible for a computer to
sample the world in much the same way we do by reading (for example)
National Geographic.
-James Rogers
jamesr@best.com