I deal with some of the worst computer systems failures worldwide. Whether it’s that or a building failure, I see that people’s minds do not operate well on low probability events. They do poorly at distinguishing a 1 in 100 event from a 1 in 10,000 event