<p>➀ Researchers developed an AI-driven method using supercomputers to predict high-performance thermoelectric materials like nickel-germanium (Ni₃Ge), accelerating discovery compared to traditional trial-and-error approaches;</p><p>➁ The AI screened thousands of metal combinations, prioritizing efficiency and reducing computational time, with experimental validation confirming Ni₃Ge's performance;</p><p>➂ The approach signals a shift in materials science strategy, with companies like Google and Microsoft actively integrating AI to enhance industrial R&D in energy and semiconductor sectors.</p>