During my undergrad days, I recall this potential problem as a critique of numerical analysis. It was stated as, "There is limited benefit of proving something

During my undergrad days, I recall this potential problem as a critique of numerical analysis. It was stated as, "There is limited benefit of proving something if we don't also know WHY it is so."