Human intuition plays a crucial role in the discovery of new quantum materials, and a new approach called Materials Expert-Artificial Intelligence (ME-AI) is bridging the gap between expert insight and machine learning. Developed by Eun-Ah Kim, the Hans A. Bethe Professor of Physics at Cornell University, and her collaborators, including Leslie Schoop from Princeton University, ME-AI captures the nuanced reasoning and intuition of human experts and translates it into a machine-readable framework. Traditional quantum materials research often relies on complex physical properties that are difficult to model using purely computational methods. While artificial intelligence can process vast datasets, it struggles to replicate the subtle, often subconscious judgments that human experts make. Kim’s team addressed this by creating a system where experts curate and define the key features of data, effectively teaching the AI to think like a human scientist. The ME-AI model was tested on a specific challenge: identifying which of 879 square-net materials exhibit desirable properties as topological semimetals. Schoop and her team provided expert-labeled data, guiding the model to focus on relevant structural and electronic features. The results showed that ME-AI not only reproduced the experts’ insights but also generalized beyond the original dataset, identifying similar promising materials in a different compound group. One of the most striking outcomes was when the model revealed a pattern that Schoop immediately recognized as her own intuitive reasoning. She said, “Oh, that makes a lot of sense,” indicating that the AI had captured her mental process. This moment highlighted a key advantage of ME-AI: while human intuition often operates too quickly to be articulated, machines can explain their conclusions step by step, making the expert’s thought process transparent and reproducible. Kim emphasizes that this approach marks a new paradigm in materials discovery—one that moves away from random or serendipitous findings toward targeted, intelligent search. The success of ME-AI hinges on high-quality, expert-curated data. Without it, even powerful AI systems can produce misleading results. The research, published in Communications Materials, lays the foundation for the National Science Foundation’s AI-Materials Institute (AI-MI), which brings together materials scientists and machine learning experts to accelerate discovery. By integrating human insight with AI’s analytical power, the team aims to unlock next-generation quantum materials with revolutionary applications in computing, energy, and beyond. As Kim notes, good data curation is essential—because in science, the quality of insight begins with the quality of data.