This paper explores the potential of artificial intelligence (AI) to accelerate advancements in physiology. It argues that while biologists often face data overload, physiologists struggle to connect genotype and structure to tissue and organ function. AI, particularly deep learning, offers potential shortcuts for predicting pathophysiological phenotypes from molecular data. However, the paper cautions that these approaches require substantial training data and may lack mechanistic explanations. A promising strategy involves using well-validated, mechanistic computational models to discover genotype-phenotype relationships and generate new hypotheses. The study focuses on cardiac myocytes and the use of machine learning to analyze isometric twitch tension, identifying features that classify disease mutations with high accuracy.