You've probably gotten a spam email or two. So how do we detect if an email is spam? And what does that have to do with classifiers, machine learning, overfitting, CAPTCHAs, the Turing test, and artificial intelligence?