S1, E6: Understanding the content of an email to know that it is spam; machine learning does so with the Naive Bayes approach; a simple application of Bayes' theorem to a simplified model of language generation; going backward from effects to their causes.