Can machine learning solve scarcity? This founder thinks so... (PLUS: Interest in MLOps "exploding" while ML research is stagnating?)