I'm aware that well-ordering of all subsets of positive integers is provable from the principle of mathematical induction, but is there a proof of the converse? I'm just curious because I've heard in a past logic course that induction depends on well-ordering.