The course addresses the question of how to explain "black-box" models that are opaque and do not provide any explanation for their inner workings. The course introduces different explainability paradigms, such as post-hoc methods, surrogate models, Shapley values, and counterfactual explanations.