Tracking Truth argued, among other things, that justified belief is not necessary for knowledge. This makes it puzzling what justified belief is and what it is good for. I am now working on a second-order tracking-based account of justified belief that displays its distinctive survival value over and above that of knowledge, and also provides a rationality constraint for coping with information about one’s own fallibility. Current papers include “Justification and the Growth of Error,” “Closure on Skepticism,” and “Optimism about the Pessimistic Induction.” Works in progress are “You can believe you’re stupid if you try” and “The Knowledge Game.”