Six US universities have launched a partnership aimed at shedding light into what happens inside artificial intelligence (AI) systems. Funded by the US National Science Foundation, the 'Understanding the World Through Code' initiative is aimed at 'turning AI black boxes into human-interpretable computer code, allowing them to solve unsolvable problems'. As part of the programme, the researchers intend to develop programming languages that can represent insights from highly complex scientific data sets. The will focus on four areas of applications: organic chemistry, cognitive science, cell biology, and engineering of complex networks systems. The involved universities are the Massachusetts Institute of Technology, California Institute of Technology, Rice University, Stanford University, University of Pennsylvania, and the University of Texas at Austin.