Diabetes Care in the Age of Algorithms

Large language models and chatbots are already part of everyday life: patients ask them to explain lab results, blood-glucose readings, meals, lifestyle or treatment. What does the diabetologist see of this in the consulting room, and what can these tools safely be used for? Modern sensors and insulin pumps analyse ever more data and support or automate certain therapeutic decisions, while other solutions - AI-based glucose prediction, meal and activity recognition, reinforcement-learning insulin dosing, digital twins - are still at the research stage. The discussion brings the researcher's, the diabetologist's and the patient's perspective to where the technology stands, and which decisions must not be handed over to algorithms.