Interesting that they study the "meaningfulness" of previously meaningless music - in a way I see a relationship to reinforcement learning, namely retroactively assigning meaning to recorded but undiscovered information from the environment.
"Information is a difference which makes a difference!"