WolframAlpha computational knowledge AI
max ln(p)ln(1-p), 0<=p<=1