Jason Yosinski sits in a small glass box at Uber's San Francisco, California, headquarters, pondering the mind of an artificial intelligence. An Uber research scientist, Yosinski is performing a kind of brain surgery on the AI running on his laptop. Like many of the AIs that will soon be powering so much of modern life, including self-driving Uber cars, Yosinski's program is a deep neural network, with an architecture loosely inspired by the brain. And like the brain, the program is hard to understand from the outside: It's a black box. This particular AI has been trained, using a vast sum of labeled images, to recognize objects as random as zebras, fire trucks, and seat belts. Could it recognize Yosinski and the reporter hovering in front of the webcam? Yosinski zooms in on one of the AI's individual computational nodes--the neurons, so to speak--to see what is prompting its response.

Deep neural networks, or deep learning, as the field is also called, have the potential to revolutionize scientific discovery. But as these networks are applied to more and more disciplines, many scientists, whose very enterprise is founded on explanation, have been left with a nagging question: Why, model, why? This interpretability problem is galvanizing a new generation of researchers in both industry and academia. Just as the microscope revealed the cell, these researchers are crafting tools that will allow insight into how neural networks make decisions. Some tools probe the artificial intelligence (AI) without penetrating it; some are alternative algorithms that can compete with neural nets, but with more transparency; and some use still more deep learning to get inside the black box. Taken together, they add up to a new discipline.