the second question is what if you overdiagnosed. what if you were sitting in this hall of mirrors, constantly being surveyed, this kind of big data is watching you. what would happen? i mean would we start intervening on cancers that we would not otherwise intervene on. maybe those cancers are harmless. so for an oncologist, for me t raises the whole question of diagnosis. you know, where are we going with diagnosis. are we going to overdiagnose, are we going to start invading in the body in a way we hadn't expected to just because as machines learn, they'll keep telling us. >> rose: help me understand the difference. i will always believe that all knowledge is good. the more you knew, the better you were. and you are saying here, if the more you know causes you to take certain kind-- too many tests. >> that's exactly right. and in fact, let me turn it back to you and say all knowledge isn't good in medicine. there are things you don't need to know about. >> rose: that's one thing, say you don't need to know about, i'm not talking a