CERN is an interesting case. They collect a tremendous amount of data, one petabyte per day. You have a lot of smart people looking for patterns in the data the reinforce or reject current thinking. Our experimental data in this case far outstrips the number of smart people we have looking at it. I would say we are in a world where the data we collect is under analysed. A single cryo electron microscope will produce 3 terabytes per day. There is stuff there we are are not seeing that our neural networks will see. New relationships between particles, new protein/cell interactions. There will be a PhD in the process for now who takes those relationships and puts theory to the test, but ten years from now maybe not.