A toy dynamical model of whether the AI workforce that builds future AI ends up cooperative or uncooperative: where the basin boundary lies, what current evidence says about which side we are on, and what would tell us we are on the good path.