The Fekete-Szeg¨o theorem gives a potential-theoretic criterion to decide if a subset of the complex plane
contains many Galois orbits of algebraic integers.

In this talk, I will explain this classical results and,
time permitting, some of its extensions and related open problems.