A human can provide guidance to the system and edit models. As to whether it “thinks like a human” I think that’s always a dangerous and loaded statement - but

A human can provide guidance to the system and edit models. As to whether it “thinks like a human” I think that’s always a dangerous and loaded statement - but what we believe is that a Bayesian probabilistic approach is closer to the way humans learn and make decisions.