An infinite number of math proofs really need more than 15M to write out. Could that include some easy-to-state ones that we care a lot about? Yeah probably.