Jeff Shallit has an interesting comparison of Stephen Wolfram and William Dembski, and their shared apparent unwillingness to admit mistakes. Over at Recursivity.