Training a neural network involves feeding it enough raw data to start recognizing and replicating patterns. It can be a long, tedious process to just approximate complex things -- like writing articles for Engadget, for example. Research scientist J...