At the end of 2020 I included in the Megatrends for this decade the use of artificial intelligence to design new alloys. There is a potential huge space for mixing atoms of different elements (plus using additive manufacturing we could create specific atomic structures!) and because of that we can dream the possibility to create an alloy with the characteristics we need. The problem is looking for a needle in a haystack.
Now I stumbled onto a news that is reinforcing that Megatrend: researchers at the University of Liverpool, UK, have developed an AI based tool aiming at reducing the time to “invent” new materials. Interestingly, their approach is based on a collaboration between material designers and AI.
They are reporting in an article published on Sept. 21st on the successful discovery of four new materials. Among these they found a solid state materials that could be applied in lithium-ion batteries to increase their performance.
The AI system is learning from existing materials (several hundred thousands) correlating their characteristic with their atomic structure and use this knowledge to explore potential new combinations that can be promising, ranking the probability. This allows material designers to focus on the most promising solutions.
I like this news because it shows a trend towards combining human with artificial intelligence in a collaborative way. Notice that here we do not have a screwdriver that is helping a human worker in a specific activity, rather we have a tool that in addition to providing help is “learning” on the way, like a team of humans where we learn from each other making the teamwork better and better. This is the crucial difference in applying AI in teamwork from applying standard tools.