If autocorrect can't work properly,then how can more complex forms of A.I. supposed to work?

Comments