4 ms·
But you're much more likely to be born into a world where this event didn't happen. The history of nuclear accidents is not a good indicator for the risk of nu
by VMG 12y ago
But you're much more likely to be born into a world where this event didn't happen.
The history of nuclear accidents is not a good indicator for the risk of nuclear war due to survivorship bias.
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.
See http://en.wikipedia.org/wiki/Doomsday_argument http://en.wikipedia.org/wiki/Doomsday_argument
- InclinedPlane 12y agoAh, interesting, a bit like a quantitative version of the anthropic principle. It presupposes that civilizations that have recovered from doomsday scenarios will never be as large as those that never had such an event occur. Perhaps on the timescale since the Cold War that holds.
- jessaustin 12y agoThis seems much sillier than the anthropic principle, however. (And I've had a professor thunder at me that the anthropic principle itself is silly, anyway.) It's not as though we know of any nuclear-capable civilizations besides our own.
- VMG 12y agoI'd love to hear an explanation why it seems silly.
- jessaustin 12y agoIn 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.