Comment by almostgotcaught

it's explicitly stated in the article:

> For an arbitrary n, compute the table full of values and their proofs, and just pull out the nth proof

if you thought harder about it you'd realize what you're suggesting is impossible