A new deep learning tool can help develop key technologies such as chips, batteries, and supercomputers.
DeepMind, a Google subsidiary, said it had discovered 2.2 million new content, which they said equates to "800 years of knowledge. According to Google's DeepMind blog, 380,000 stable materials have been found that can be used in future technologies, including a new generation of computer chips and batteries, as well as solar panels. The discovery was made using the advanced GNoME neural network tool (Graph Networks for Materials Exploration). The results of the study are published in the prestigious scientific journal Nature.
DeepMind's Amil Merchant and Ekin Dogus Cubuk noted in a blog post: "With GNoME, we have increased the number of technologically viable materials known to humanity. Among the candidates found are materials that can form the basis for the development of future transformational technologies, including superconductors for supercomputers and next-generation batteries for improving the efficiency of electric vehicles."
They also pointed out that using GNoME allowed them to bypass the centuries of "painstaking experiments" required to discover new materials.
"Using GNoME shows how powerful artificial intelligence can be in finding and creating new content... We believe that GNoME and other AI tools can fundamentally change the process of opening materials and determine the future of this industry," the blog notes.
It is noted that external researchers verified the results of DeepMind's work, creating 736 new materials opened using GNoME.
DeepMind is an artificial intelligence technology that uses machine learning to solve problems that computers traditionally couldn't handle. Its technology is spreading rapidly and is currently being used in several real-world applications.
The company was founded in London in 2010 and acquired by Google in 2014. It is now a subsidiary of Alphabet Inc, the parent company of Google.
DeepMind, a Google subsidiary, said it had discovered 2.2 million new content, which they said equates to "800 years of knowledge. According to Google's DeepMind blog, 380,000 stable materials have been found that can be used in future technologies, including a new generation of computer chips and batteries, as well as solar panels. The discovery was made using the advanced GNoME neural network tool (Graph Networks for Materials Exploration). The results of the study are published in the prestigious scientific journal Nature.
DeepMind's Amil Merchant and Ekin Dogus Cubuk noted in a blog post: "With GNoME, we have increased the number of technologically viable materials known to humanity. Among the candidates found are materials that can form the basis for the development of future transformational technologies, including superconductors for supercomputers and next-generation batteries for improving the efficiency of electric vehicles."
They also pointed out that using GNoME allowed them to bypass the centuries of "painstaking experiments" required to discover new materials.
"Using GNoME shows how powerful artificial intelligence can be in finding and creating new content... We believe that GNoME and other AI tools can fundamentally change the process of opening materials and determine the future of this industry," the blog notes.
It is noted that external researchers verified the results of DeepMind's work, creating 736 new materials opened using GNoME.
DeepMind is an artificial intelligence technology that uses machine learning to solve problems that computers traditionally couldn't handle. Its technology is spreading rapidly and is currently being used in several real-world applications.
The company was founded in London in 2010 and acquired by Google in 2014. It is now a subsidiary of Alphabet Inc, the parent company of Google.
