MrMind comments on Patternist friendly AI risk