I read in sci.electronics.design that John Larkin wrote (in ) about 'failure', on Fri,
30 Sep 2005:
Because there is no way in which it could do so. It would need cognition in order to comprehend the form that a new gene should take in order to produce proteins that would in some way counteract the threat.
Unless, of course, it produced a few hundred thousand proteins at random and only the organisms that produced the right ones survived. Hey, that's a neat idea. We could call it 'evolution by natural selection'. Probably wouldn't catch on, though. Too much opposition from religious quarters.
Occam. There is no need to postulate (but not define!) 'very complex meta-mechanisms' when a simpler postulate explains all the evidence.
Occam. He's got a razor.