In deference to my professor, I'll just observe why "The fact that we haven't observed a nuclear war in history might have less to do with the likelihood of nuclear war and more with the fact that fewer people would be able to observe that history," seems sillier than Carter's (not Tipler's!) anthropic principle. The anthropic principle considers only two classes of universe: one that will at some point host observers of any sort, and the other that will not. Clearly we find ourselves in a universe of the former class. Whereas your statement invokes any number of possible histories and events, and blithely assigns probability distributions to them. I value parsimony.
