Novel functional, or "smart," materials are revolutionizing nearly every facet of scientific and industrial processes. Multidisciplinary teams of researchers at the Weizmann Institute have helped discover insights into the fundamental behavior of atoms, molecules, and crystals with implications in wide areas of research, from chemical and biological reactions to materials science and condensed-matter physics. Our research has contributed to the advancement of applications ranging from bioengineering to renewable energy sources.
However, the optimization of novel applications often involves costly and time-consuming tests of the relevant key parameters. Streamlining the development of these materials into a quicker, more efficient, and more accessible process requires understanding the relationships between structure, order, and the dimensionality of the atomic components in each novel molecular design, and how they influence and enable desired functions.
AI generative models can pave the way for systematic analyses of large datasets, potentially leading to the discovery of new design rules. This necessitates overcoming challenges such as creating new tools for collecting these extensive datasets, and developing advanced methods to glean new understanding from the data, like constructing "explainable AI" tools that not only provide conclusions but also explain the reasoning behind them. These models also require powerful computational architectures to support their complexity and facilitate the development of new materials for various fields, ranging from bioactive molecules for therapeutics and other purposes to optimal energy harvesting and quantum information transfer.