Researchers at Paige and Microsoft have developed PRISM2, an AI model that interprets pathology slides by integrating visual information with clinical dialogue from pathology reports. This approach enables the model to generate detailed text-based answers to diagnostic questions, moving beyond simple pixel classification. The model's training data consists of a vast collection of whole-slide images and corresponding clinical text. By combining these two sources of information, PRISM2 aims to provide more accurate and informative diagnostic outputs. This advancement has the potential to improve the efficiency and accuracy of medical diagnoses.