AI Ethics: Should you trust AI with your medical diagnosis?

Abstract

The use of AI in the medical field leads to one main ethical question: Should you trust AI with your medical diagnosis? To answer this question, one must understand how AI decides, which faces two main problems: The black box and validation data. The first is a problem of transparency in which we have no idea how AI makes decisions. The second is a problem of how we can set the validation data for training AI. This paper aims to analyze both problems and clarify how AI decides in medical diagnoses. Then I will show that AI use for medical image processing is one of the models that doesn't face these problems since it can provide the evidence of diagnosis. Next, I will address the question of whether you should trust AI in medical diagnoses, and I will show that the answer depends on comparing the functions between humans and AI.

Keywords

• ai ethics
• ai in medical diagnosis
• medical image processing
• consequentialism