It's often said that evolution works in such a way as to maximize the number of descendants that an individual has.
More formally, X's children share one-half of their genes with X, X's grandchildren share one-quarter of their genes with X, and more generally X's nth-generation descendants share 1/2n of their genes with X. So if you buy the whole "selfish gene" theory that genes act in such a way as to maximize the number of copies of them which are made, the quantity individuals should be attempting to maximize might be half the number of children, plus one fourth the number of grandchildren, plus one eighth the number of great-grandchildren, and so on.
It's an infinite sum.
What's more, if you assume a "total fertility rate" of k -- that is, the average female bears two children -- then this sum is k/2 + k2/4 + k3/8 + .... And if k = 2, which corresponds to population not growing or shrinking, this sum is just 1 + 1 + 1 + 1 + ... which doesn't converge. (Similarly if k > 2, but populations which grow indefinitely don't seem sustainable.)
Of course, biologically populations don't live for infinitely long. And in reality people don't at least consciously think more than a couple generations into the future. So practically speaking this is all a bit meaningless.
edited, Monday, 8:42 am: I did mean maximizing the number of copies of genes, in the sense of Dawkins' idea of the selfish gene, and this post is not meant to be taken seriously.