Google's AI unit Deepmind says it discovers '800 years' worth of knowledge

Google's AI division, DeepMind, today said it has achieved a groundbreaking feat in material science with its AI tool, Graph Networks for Materials Exploration (GNoME).

According to Google Deepmind, the tool has identified an unprecedented 2.2 million new crystalline materials, including 380,000 stable ones, which could significantly advance various modern technologies.

What is the importance of stable crystals in technology?

Stable inorganic crystals are foundational in developing technologies like computer chips, batteries, and solar panels. GNoME's discovery of stable materials is crucial as these crystals must not decompose for technological viability. According to Deepmind, this could lead to transformative technologies in superconductors, advanced batteries, and more efficient electric vehicles.

"With GNoME, we've multiplied the number of technologically viable materials known to humanity. Among these candidates are materials that have the potential to develop future transformative technologies ranging from superconductors, powering supercomputers, and next-generation batteries to boost the efficiency of electric vehicles," Amil Merchant and Ekin Dogus Cubuk said in the blog post.

GNoME's advanced AI capabilities

GNoME represents a significant leap in AI-driven material science. It's a sophisticated graph neural network model, trained with data from the Materials Project. GNoME excels in predicting material stability, a task that was traditionally slow and costly. It has dramatically increased the speed and accuracy of material discovery, with a prediction success rate of about 80%, the blog post said.

Global collaboration and independent verification

DeepMind's discovery has garnered global attention, with external researchers independently synthesizing 736 of these new materials. Collaborations, like with the Lawrence Berkeley National Laboratory, have further validated GNoME's predictions through autonomous material synthesis, highlighting the practical applicability and global impact of DeepMind's AI research.

DeepMind said it is democratizing access to the wealth of knowledge by making GNoME's predictions public and contributing the findings to the Materials Project database, enhancing global research efforts in inorganic crystals. This move is expected to catalyze further research and development in the field, leveraging AI as a tool for experimental guidance.

Join our new WhatsApp Channel for the latest startup news updates