Key points are not available for this paper at this time.
The evolution of materials science is undergoing a profound paradigm shift driven by artificial intelligence (AI), transitioning from traditional intuition-driven trial-and-error to an accelerated, data-centric, and algorithmically guided discipline. This review examines this transformation through the lens of the materials discovery workflow, structured around two parallel and complementary trajectories. First, we discuss task-specific AI for materials science. We detail its role in distinct stages of the materials discovery pipeline, including hypothesis generation, experimental planning and optimization, characterization, and knowledge discovery. Second, we explore generalist AI for materials science, designed to handle universal scientific tasks. We examine how these systems advance knowledge representation, enable agentic workflows that orchestrate autonomous laboratories, and facilitate human-AI collaborative reasoning. Finally, we provide perspectives on the future ecosystem of AI for materials science (AI4Mat), outlining the critical challenges and strategic directions that must be addressed to realize the full potential of this evolving discipline.
Li et al. (Wed,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: