Researchers at the Karlsruhe Institute of Technology (KIT), Germany in collaboration with partners, are demonstrating how artificial intelligence (AI) can help identify new research ideas in materials science. In a recent study, they analyse large volumes of scientific literature using a combination of large language models and machine learning. This allows trends, previously overlooked connections and potential new research approaches to be systematically identified.

Given the rapidly growing number of publications, even experts are finding it increasingly difficult to keep track of everything. “Our aim is to support researchers in creative thinking processes by highlighting new research questions and potential collaborations between disciplines,” says Professor Pascal Friederich from KIT.

The work underscores the importance of materials science as an interdisciplinary field spanning technologies from energy solutions to medical technology.

The study has been published in Nature Machine Intelligence: Marwitz, T., Colsmann, A., Breitung, B. et al. Predicting new research directions in materials science using large language models and concept graphs. Nat Mach Intell (2026). https://doi.org/10.1038/s42256-026-01206-y