The earliest studies on necessary and sufficient neural populations were performed on simple invertebrate circuits. Does this logic still serve us as we tackle more sophisticated outputs?