Lecture: Putting Turing to Work

Alan Kay

Abstract:

Many computer programs today are enormously large -- 10s and 100s of millions of lines of code -- and distressingly messy and buggy -- many to the point that programmers are afraid to make major revisions, but must resort to adding still more code "around the outside". Though it is hard to prove interesting and useful things about computations, analogies to some of the expressive ideas in Mathematics can be used to make programs much more expressive and much smaller. Similarly, though the scales are not commensurate to really use some of the deep ideas from Biology, many of the architectural principles that produce life from inanimate matter can be adapted to make large systems which "find ways to work and keep working".