What do Aristotle, the medieval Scholastics, and individuals such as Descartes,
for example, have to do with computers and AI? Well, it turns out that until
the change of world-view that came with the Scientific Revolution at the end
of the seventeenth century, the very notion of a machine being capable of reproducing
functions of perception and cognition was literally unthinkable. As another
example, most people with an interest in this kind of thing are aware of Charles
Babbage and his design for a steam-driven mechanical computer, the "Analytical
Engine," which was years ahead of its time but never built because of a
combination of technical and political problems. But how many realize that he
also designed a "Difference Engine" that was built, worked very well,
and does so to this day for that matter? We not only go into that, but also
what it did and why there was such a desperate need for it.
Other topics that we cover include:
-- logic and the story of attempts to mechanize it; how Russell and
Whitehead torpedoed Gottlob Frege when he thought he could prove all of mathematics,
only to be blown out of the water themselves by Kurt Gödel; Alan Turing with
his machines and his Test; what set the digital computer apart from every
other machine ever conceived.
-- the development of AI from the "Cybernetics" of the 40s,
the Dartmouth Conference in 1956, through to the work in the 60s and 70s at
places like MIT, Carnegie Mellon, Stanford, and IBM on such things as theorem
proving, problem solving, artificial vision, robotics.
-- computer game-playing from Arthur Samuel's Checkers Player in the
50s to IBM's Deep Blue.
-- the enormous difficulties of trying to program such faculties as
natural-language comprehension and common sense; why computers that are so
good at the things humans find hard have so much trouble with tasks that are
effortless for young children.
and along the way, a look at such techniques as expert systems, parallel
architecture, neural nets and Boltzmann machines, data mining, genetic programming,
and holographic processing.