1: \begin{abstract}

2: We describe the analogue of the Sato-Tate conjecture for an abelian variety over a number field; this predicts that the zeta functions of the reductions over various finite fields, when properly normalized, have a limiting distribution predicted by a certain group-theoretic construction related to Hodge theory, Galois images, and endomorphisms. 

3: After making precise the definition of the \emph{Sato-Tate group} appearing in this conjecture, we describe the classification of Sato-Tate groups of abelian surfaces due to Fit\'e--Kedlaya--Rotger--Sutherland.

4: \end{abstract}