But how do you know which problems are worth making progress on? Traditionally, you'd find those by working on solving previous problems.