"Taking `Hard' Problems to the Limit," by Barry Cipra. Science, 14 March 1997, page 1570.
There are some problems in theoretical computer science that seem to be extremely hard to solve; so hard, in fact, that despite great efforts no one has come up with efficient algorithms for solving them on a computer. Are these problems inherently hard, or is it just that no one has been clever enough to hit upon the right algorithm? This article describes the highly speculative ideas of Michael Freedman, a world-famous topologist, for a way to attack this question. Freedman's idea centers on a careful examination of what happens when the size of these problems grows infinitely large.
--- Allyn Jackson