Deepmind’s GNoME AI managed to create 2.2 million materials, the most stable of which could prove revolutionary. Some of them have already been synthesized.
This will also interest you
[EN VIDÉO] Physics: Plasma could revolutionize industrial production Matter is solid, liquid or gaseous, but there is a fourth state that we…
Artificial intelligence is causing a major upswing in many areas and Google’s AI, Google Deepmind, has proven this once again. It has just saved 800 years of science by creating new materials with revolutionary potential. The tool Graph Networks for Materials Exploration (GNoME) from Deepmind is responsible. Its algorithms discovered 2.2 million new inorganic crystals and identified 380,000 of them as the most stable. From this list of stable materials, researchers began to synthesize them. Without this AI, it would have been necessary to conduct endless experiments and tinkering in the creation of these materials. This push from GNoME has already helped synthesize 736 of them in research labs around the world.
An autonomous laboratory for the synthesis of materials
These candidates include materials that could revolutionize superconductors, supercomputers and batteries to increase the efficiency of electric vehicles. These discoveries include 52,000 new graphene-like compounds. Another example: 528 potential lithium-ion conductors that could improve battery performance. The Deepmind team also worked with Berkeley Lab to create a laboratory for the automated and autonomous synthesis of these new crystals. The robots in this laboratory have already managed to create 41 of these new materials.