WolframAlpha computational knowledge AI
max (-x2-x*M+x)/(1-x) with 0<x<1 and 0< M<1