less than 1%. it's really important and it's not just medical scans and slides and all the things that we are looking at with the data, we can get augmented help from machine algorithms. >> you make this point often that it's not that the machines will replace the human beings, that they'll augment them. when i hear that business about the scans i think to myself, why do you need a radiologist if the machine is doing a better job? >> for lots of reasons. because you don't want to entrust a machine with a human life. moreover, the machine has no contextule capability. all it can do is say i saw this nodule. s where the human factor which is the oversight that backup is so important to put it in perspective for that patient. >> correct me in i'm wrong, but you see the machines doing a lot of what we train doctors to do now kind of highly and littic, pattern recognition but you're saying the machines can do that better than the the humans.