Edward Hirst, Henrique N. Sá Earp & Tomás S. R. Silva

Brief Description

“AI meets Geometry” is a hands-on minicourse introducing modern artificial intelligence as a practical tool for exploration, computation, and discovery in differential geometry. Participants will learn how neural networks, physics-informed methods, optimization, and reinforcement learning can be used to approximate geometric structures, study metrics and differential forms, investigate geometric flows, and search for new examples and conjectures, while keeping clear the essential distinction between numerical discovery and mathematical verification. No prior machine-learning background is assumed: the course builds conceptual foundations alongside concrete geometric case studies, including special metrics, neural PDE methods, and AI-assisted discovery. Code will be provided throughout, giving attendees working notebooks that they can modify, extend, and adapt directly to questions arising in their own research.