Robert Hart, Cambridge graduate, discussed a very interesting, yet becoming increasingly important issue:
WHO is to blame if a robot AI system makes a misdiagnosis?
The current state of the art is that medical doctors are using AI today as decision support, definitely not as a replacement for standard procedures. However, what if technological progress keeps on and AI gets better and will be directly included in decision making systems. Ist the doctor then still be responsible for any error that might occur and result definitely from AI?
This is a very hard question but it will need answers in the future. For example, Geoffrey Hinton, pioneer in Deep Learning, already emphasizes that e.g. radiologists will be replaced by deep learning approaches (see a discussion on Geoffrey Hintons provocation here: https://medium.com/unauthorized-storytelling/the-radiologist-is-here-to-stay-24da650621b5
However, whether and to what extent it will come or not, the question remains open: Who is then responsible if an error occurs?
Read the original story here:
https://qz.com/989137/when-a-robot-ai-doctor-misdiagnoses-you-whos-to-blame