An international physics consortium is using machine learning to find room-temperature superconductors — and the search just got dramatically faster.
Picture a power cable that wastes absolutely nothing. No heat, no lost energy, no hum of resistance — just electricity flowing as if friction had been uninvented. That is what a superconductor does. The catch?
Every superconductor we know of needs to be cooled to temperatures close to absolute zero before the quantum magic begins. Now, a global research team thinks artificial intelligence can help crack the problem wide open. The SuperC consortium, founded in 2023 by Aalto University's Professor Päivi Törmä, is the first coordinated worldwide collaboration with a single, audacious target: discover a room-temperature superconductor by 2033. The team's latest results show that machine learning can sweep through an almost infinite landscape of chemical combinations and pinpoint the most promising candidates in a fraction of the time it would take any human researcher.
Two new materials — YRu3B2 and LuRu3B2 — have already been identified this way.