Princeton University researchers Arvind Narayanan, Aylin Caliskan and Joanna Bryson discuss their research on how human biases seep into artificial intelligence.