Dentin comments on Patternist friendly AI risk