Frontiers of Materials That Learn

New advances in materials research are unlocking the potential of materials that learn. These materials - whether chemical networks that "decide" on their own, biohybrid materials, or tiny robot swarms - can have the ability to sense and adapt to new environments, perform intelligent tasks such as image recognition, and serve a structural purpose. These physical systems could consume less power and prove more resilient to damage than traditional electronic materials, making them particularly useful for extreme conditions such as high radiation or aqueous environments.

These proceedings summarize an October 2025 workshop that brought together researchers from across materials design, condensed matter physics, biomaterials, and industry to discuss new opportunities for materials that learn and ways to move towards the scalable design and manufacturing of these learning systems.