Well, of course I didn't mean to suggest p=0. I don't think the collapse of technological civilization is very likely, though - and would assign permanent setbacks a < 1% chance of happening.