Large language models like GPT show that machines can now understand text and continue stories. But can similar model architectures also grasp the story of a field – with weather, soil, crops, and machines as key players? This talk provides an overview of current research approaches and explores how AI could, in the future, help predict and shape agricultural processes.