methods Context-adaptive systems Training models to reveal heterogeneity by automatically adapting to context Foundation models as context Bridging foundation models and statistical inference Intelligible ML Building glass-box models that are both predictive and interpretable Automated data science Automating data science with foundation models applications Genomics Dissecting the biological bases of disease Clinical AI Tools Building tools for practicing clinicians Real-world evidence Building AI systems to reveal surprises in observational EHRs Maternal health Applications to understand risk factors underlying maternal morbidity.