They further explained that exploring every possible electrolyte combination using lab testing alone would be nearly impossible, with the number of potential candidate molecules estimated to go as high as 10⁶⁰.
“It would have been impossible for us to go through hundreds of millions of compounds to say, ‘Oh, I think we should study this one,'” Amanchukwu stated, and compared using AI in research to listening to music online.
Like music recommendation algorithms, the AI learns patterns to predict the most promising molecules. Amanchukwu stated that the next step is developing an AI that doesn’t just pick songs, but also creates playlists, thus paving the way for the design of entirely new molecules.