The recipe for our favorite dish is perfected through trial and error. This is also true for research in the field of molecular studies.
A team from the Technical University of Munich employs an artificial intelligence that seeks out usable data on its own.
While developing, say, organic semiconductors for solar cells, the team must first calculate the conductivity of a small set of molecules, data which is then examined by the AI.
After sifting the data, the AI picks out the most promising and develops it further. Learning as it goes and becoming more efficient.
The technology makes the process of creating an ideal material a lot more streamlined by reducing the amount of testing.
Quality over quantity.