WolframAlpha computational knowledge AI
max {p^2+4 (1-p) p}