By analogy, if I write a computer program to perform some calculation, I don't write it so that every now and then I have to stop the program and rewrite the code in order to get it to do what I want. I'd be a pretty lame programmer if I did. Instead I write it so that I press the start button and the program does the rest.
I have to disagree here. You'd be an pretty extaordinary programmer if you only wrote programs which never needed any fixes or updates. And along the same lines this arguement:
The history of evolution is littered with failures and dead-ends. We can point to species that didn't make it ... and genera ... and families ... and orders ... and classes ... and phyla. At the lowest level, we see lots of failed mutations that will never make it: mutations that lead to death or severe handicap or sterility. Does this look like the product of a perfect, all-knowing God, or of the hit-and-miss processes described by the theory of evolution?
fails in that it assumes god is either perfect and all-knowing or non-existant. If we assume that a god exists, I think the evidance is pretty ample for imperfection.
That being said, I'd say the only arguement possible for guided evolution is "I have faith it is so, and you can't disprove faith"