In American today, it is impossible to be an atheist and not have some opinion, well-founded or otherwise, on Intelligent Design. Of course, it's a given that an infidel atheist would be deeply suspicious, if not outright hostile to the hypothesis that there is a guiding hand at work in descent with modification -- that's how Darwin referred to "evolution."
At the bottom of this suspicion is an intuition that something just isn't right with Intelligent Design. But until I read "What is wrong with Intelligent Design?" by Elliott Sober, appearing in the March 2007 issue of The Quarterly Review of Biology, I couldn't quite put my finger on the problem.
A synopsis of the article can be found here.
The real meat of the article is an elegant and surprisingly simple explanation of why Intelligent Design isn't science. The choice between competing theories is made by examining the predictions that follow from the possible theories. These predictions are verified by independent evidence. A theory of eclipses that uses gravity and a knowledge of the moon's orbit about the earth and the earth's orbit about the sun will allow for predictions that can be independently verified using optics.
However, when an Intelligent Designer is used to explain the complexity of the eye, there are no predictions that can be made that follow from postulating the existence of an Intelligent Designer. Consequently, whatever the issues surround the existence of an Intelligent Designer might be, they are not within the purview of science.
Although Sober doesn't point this out, the hypothesis of an Intelligent Designer is superfluous. Positing an Intelligent Designer to explain biological phenomena is analogous to prefixing "I posit that" to every statement of fact that a person might make. "I posit that the sky is blue." "I posit that ..."