At the 1st NC Biotech Innovation & Investment Summit, Duke University Distinguished Professor Bruce Donald delivered a compelling talk on using AI constrained by physics and chemistry — not just data — to predict cancer mutations and design drugs in advance. He highlighted fundamental limitations of current AI models in molecular biology, emphasized the importance of explainable models over black-box predictions, and introduced his paradigm-shifting approach of treating proteins as probability distributions rather than static structures. Dr. Donald, founder of Ten63 Therapeutics, demonstrated how AI and physics-based molecular design platforms are tackling challenging drug targets in oncology.