Standards in Artificial Intelligence

rick++ rickotto at
Fri Nov 3 11:05:00 EST 2000

I call this the "fallacy of the hammer" from the saying "if you give
someone a hammer, the whole world looks like a nail".

This fallacy is molding the problem to fit the tool (e.g. "64"-bit
rather than finding the best methods that solve the problem.
In computer science the tools are so transitory; The hardware and
ten years ago was rather different, and could be ten years from now.

We've seen a couple generations of the hammer fallacy in A.I. including
Sussman's LISP machines and Hillis's Connection Machine without
commesurate progress in A.I.

Scientists and engineers aren't immune to this either, worrying too
much about their next "super-duper-computer" than the science.

Sent via
Before you buy.

More information about the Neur-sci mailing list