Is there a polynomial time algorithm for computing the Reshetikhin-Turaev invariant of a 3-manifold at the 4th root of unity? Probably not, but trying to find one leads to some pleasant and fairly easy math.