Cloud Hosting

HUNDREDS of thousands of new materials have been discovered by scientists, thanks to the help of AI-driven robots.

Researchers with Googles artificial intelligence research lab DeepMind have recently uncovered over 2.2 million new types of crystals including 380,000 materials that are stable.

Two papers published in Nature detail the findings from the researchers and scientists from DeepMind and the University of California, Berkeley.

Not only did the researchers celebrate that the finding is equivalent to nearly 800 years worth of knowledge, they indicated that their discovery can have a strong impact on future technological advances, particularly when uncovering new materials.

Specifically, they claim that among the 380,000 most stable materials discovered, they have the potential to be used to power superconductors, supercomputers, and even batteries in electric vehicles.

This discovery was largely made possible by Deepminds Graph Networks for Materials Exploration (GNoME) a deep learning AI tool.

In order to uncover millions of materials, GNoME was supposedly trained with information on various crystal structures and their stability, from data supplied by the Materials Project.

From there, the scientists had robots driven by the GNoME AI generate novel candidate crystals and predict their stability essentially cooking up new materials.

These robotic chefs quickly went to work, successfully synthesizing 41 materials out of 58 within 17 days in their "kitchen" at A-Lab, a facility at Berkley Lab.

"We now have the capability to rapidly make these new materials we come up with computationally,Berkeley materials scientists and A-Lab leader Gerbrand Ceder told Nature.com.

Ceder insisted that this technology and Ai-driven recipes for new materials will "change the world."

Not A-Lab itself, but the knowledge and information that it generates," he said.

Within this research, the scientists also found that more than 500 of the stable material candidates had "promising" lithium-ion conductors, which is a critical component in batteries.

Not only did the researchers say they would make their database of the uncovered materials available to the scientific community, they also plan on providing the recipes developed by GNoME for further testing.

"This is really exciting, Alexander Ganose, a materials chemist at Imperial College London, told Science.org.

It is enabling materials discovery across a much wider composition range. We might be able to find the materials of the future in this data set.

Google DeeMind has celebrated its findings as a large advancement of using AI technology in scientific research.

Our research and that of collaborators at the Berkeley Lab, Google Research, and teams around the world shows the potential to use AI to guide materials discovery, experimentation, and synthesis, researchers Amil Merchant and Ekin Dogus Cubuk said.

We hope that GNoME together with other AI tools can help revolutionize materials discovery today and shape the future of the field.

View post:
Google DeepMind AI discovers 380,000 new materials unknown to humanity and a robot chemist is already mak... - The US Sun