Do scientists have a right to be wrong? Does state-of-the-art science really enable us to speak confidentially about “low-probability” events? And is there a way to better communicate risk to the general public?