A diagnostics and digital pathology advance focused on how foundation models can be engineered for robustness in clinical microscopy workflows. Researchers described a foundation-model approach for digital pathology designed to improve consistency, accuracy, and adaptability—aiming to reduce common deployment risks where models can degrade across scanners, stains, and clinical sites. In computational drug development, a new AI modeling approach was described as fast enough to accelerate molecular simulation trajectories by orders of magnitude, potentially enabling quicker identification and prioritization of candidate molecules. The reported capability could shorten the feedback cycle between hypothesis generation and experimental testing for future drug discovery programs.