Watch: NAO Robots Display a Brief Moment of Self-Awareness
These days artificial intelligence is an undeniably a popular subject in both Hollywood and the world of technology. But when it comes to science fiction, the concept of artificial intelligence has been popular from the very start, with the genre often exploring the idea of whether or not artificially intelligent robots can exhibit self-awareness. Now, as technological advances are made in the real world, the topic has once again moved back into the public consciousness.
Innovatively designed robots like the adorable Pepper and NAO, created by the Japanese company Softbank Group, are recent examples of robots that can interact with their owners and, in Pepper's case, even recognize human emotions. Recently, Selmer Bringsjord, a professor at the Rensselaer Polytechnic Institute who specializes in the logio-mathematical and philosophical foundations involved in artificial intelligence, gave a trio of NAO robots a wise-men test in order to see if they could exhibit self-aware behavior. The test was a variation on The King's Wise Men test; a logic puzzle solved by applying the principle of induction, and it produced some rather remarkable results.
Bringsjord programmed the robots to believe that two of the three had been giving a "dumbing pill" (not an actual pill but the push of a button on the top of their heads) to silence them. He then asked the robots to determine which two of them had received a dumbing pill, a question to which one responded, "I don't know". Upon hearing its own voice it realized it had not been giving a dumbing pill and changed its answer accordingly.
Because the NAO robot was able to recognize its own voice and respond to the question, the results of this experiment indicate that the robots are able to display at least some small degree of self-awareness. Although current technologies are nowhere near achieving human levels of self-awareness, the results of the experiment are still encouraging as robots continue to revolutionize the world as we know it.
"Soon enough, much of what many humans do for a living will be better done by indefatigable machines who require not a cent in pay," says Bringsjord. "I figure the ultimate growth industry will be building smarter and smarter such machines on the one hand, and philosophizing about whether they are truly conscious and free on the other."
The experiment run by Bringsjord will be presented at the RO-MAN 2015 conference in Japan later this month.