Artificial intelligence has become one of the key technologies in medicine, and becomes more important every year. Deep learning is one of the most important AI methods, particularly when dealing with images or movies. In this talk I will introduce deep learning, with a particular focus on people who are expert in a biomedical area but not in AI. I will discuss how it works, at a nontechnical but functional level, the differences between traditional computational approaches and deep learning, how networks are trained, and which areas they are ineffective at, excel at, and will hopefully excel at in the future. The aim is to help you decide whether deep learning could help you, and if so how to start.