This piece provides a great illustration of the importance of having good data & lots of it. We see weird predictions like water as a solvent happen when screening thousands of candidates with data-limited models. Fortunately these can be easily ruled out by intuition. 3/n
From the time we published our first #ML/#AI for battery design paper in 2017 (https://t.co/IFDOtJRIjc), it was clear this would change how R&D is done. The central question now is how to do it most effectively, and what outcomes to expect. 2/n
This resonated with me because it also defines our strategy at Aionics. We are often how we are different from others in the battery/ML space. Well - you have to ship to compete, and and Aionics has been shipping all kinds of battery AI solutions for the last three years. 4/