Fei-Fei Li, a Chinese-American expert in deep learning and a professor at Stanford University, says how AI is engineered, and by whom, will determine whether it helps all humanity or reinforces the wealth divide and human prejudices.