From Max Roser, responding to Alex.  To be clear, I don't think you are supposed to take these as highly exact estimates with respect to event independence and the like.  It is simply one way of showing that you need to think about cumulative risk.