Abstract

AI needs to develop more solid assumptions, falsifiable hypotheses, and rigorous experimentation.