from "Jeremy Avigad - Computers in Mathematical Inquiry"
The challenge of developing theories of rationality that do not assume logical omniscience are not limited to modeling mathematical beliefs; it is just that the difficulties involved in doing so are most salient in mathematical settings.