DeepMind AI Breakthrough , Could Revolutionize , Battery and Chip Development .<br />'Time' reports that researchers at Google <br />DeepMind have used artificial intelligence <br />to predict structures of 2.2 million new materials. .<br />'Time' reports that researchers at Google <br />DeepMind have used artificial intelligence <br />to predict structures of 2.2 million new materials. .<br />DeepMind published 381,000 of the <br />crystal structures that the AI predicts <br />to be the most stable.<br />The breakthrough, which could have wide-reaching <br />benefits for renewable energy and computing, increases <br />the number of known stable materials by a factor of ten.<br />While many of the materials still need to be <br />synthesized and tested, the findings are <br />predicted to accelerate the discovery of new materials. .<br />Those new materials could play <br />a central role in energy storage, <br />solar cells and superconductor technology. .<br />While materials play a very critical <br />role in almost any technology, we as <br />humanity know only about a few <br />tens of thousands of stable materials, Ekin Dogus Cubuk, a Staff Research Scientist at Google Brain, via 'Time'.<br />'Time' reports that only specific combinations <br />of elements can be used to form stable solids.<br />Google DeepMind has now partnered with <br />external researchers who have successfully <br />synthesized 736 of the new materials.<br />Among the 381,000 materials added to the Inorganic <br />Crystal Structures Database (ICSD) were 528 potential <br />lithium ion conductors that could be used in batteries.<br />52,000 new layered compounds were also predicted with <br />a similar structure to graphene, which could be <br />used to produce new superconducting materials. .<br />We believe that some of <br />these will be made in the lab, <br />which will hopefully lead <br />to very exciting applications, Ekin Dogus Cubuk, a Staff Research Scientist at Google Brain, via 'Time'
