AI offers a powerful opportunity to accelerate the discovery of chemicals and materials critical for energy and sustainability, much as it has transformed other domains. Progress in AI increasingly appears to depend on scaling—perhaps above all else. As Rich Sutton’s “bitter lesson” suggests, methods that effectively harness computation tend to outperform those relying heavily on handcrafted knowledge. In chemistry and materials science, this insight carries both promise and tension.

This paradigm is evident in the success of machine learning interatomic potentials (MLIPs), surrogates for quantum mechanical calculations. Trained on massive datasets—often containing hundreds of millions of samples—these models exhibit scaling behavior similar to that seen in language and vision systems. They generalize to new scientific problems and enable simulations that were previously infeasible. Generative models, often trained on similar synthetic data, can dream new materials. Meanwhile, debate continues on the role of inductive bias—whether physical constraints are best embedded in model architectures or learned through training.

Yet MLIPs remain proxies for simulation, and materials ultimately matter only when synthesized and deployed at scale. This is the “bitter” reality: translating AI-driven discoveries into real-world impact requires slow, costly experimental and industrial processes. Our work addresses both extremes—scalable ML-driven simulation and the hands-on challenge of turning AI-designed materials into practical technologies.

Rafael Gomez-Bombarelli is the Paul M Cook Associate Professor in MIT’s Department of Materials Science and Engineering; and co-founder and Chief Scientific Officer at Lila Sciences. He earned his undergraduate and graduate degrees in chemistry from the University of Salamanca, Spain, and conducted postdoctoral research at Heriot-Watt University and Harvard University, under Alan Aspuru-Guzik. Rafa’s research integrates machine learning with atomistic simulations to expedite the discovery and design of novel materials for applications in energy, sustainability, and healthcare.