'In 1949 neurophysiologist Grey Walter built two robot "tortoises" to show that complexity could arise out a very simple nervous system. "Elmer" and "Elsie" each had a light sensor, a touch sensor, a propulsion motor, a steering motor, and two electronic valve-based "neurons."
He found that even with this modest equipment they were capable of phototaxis, finding their way to a recharging station when their batteries ran low. In a subsequent experiment he watched as a robot moved in front of a mirror and responded to its own reflection. "It began flickering," he wrote. "Twittering, and jigging like a clumsy Narcissus." He argued that if this behavior were observed in an animal it "might be accepted as evidence of some degree of self-awareness.'