The following paradox: let n be the largest integer; if n > 1, then n² > n, which contradicts n being the largest integer and so n = 1 is the largest integer. The ridiculous conclusion shows the danger of simply assuming a solution to an optimization problem exists.