The King’s Wise Men is the simplest induction puzzle there is: The King called the three wisest men in the country to his court to decide who would become his new adviser. He placed a hat on each of their heads, such that each wise man could see all of the other hats, but none of them could see their own. Each hat was either white or blue. The king informed the wise men that at least one of them was wearing a blue hat, thus there could be one, two, or three blue hats, but not zero. The king also promised that the contest would be fair to everyone of the three men. The wise men were not allowed to communicate with one another. The first one of the three men to stand up and answer correctly what color is the hat on his head would become the King’s new adviser. After awhile, one man stood up and announced the answer. The answer was correct: His hat was blue … and so were the two others.
A variant of this logical puzzle, simplified and adapted to robotics, was used by Rensselaer Polytechnic Institute professor Selmer Bringsjord in an AI experiment.
The three cute and polite NAO robots (and not the latest model at that) were made aware of a “dumbing pill”. What’s that? It’s just a button on top of their heads. Tap the button and robot will be silenced — go “dumb.” The tester taps the head of each robot but silences only two out of the three. One is able to speak. Neither knows whether or not it was silenced. The robots then asked which one “received the dumbing pill.”
All three of the robots will try to answer “I don’t know.” The one that wasn’t silenced stands up and speaks up, “I don’t know.” It’ll take a robot only a second to realize that since it can speak, it was the one who didn’t receive the dumbing pill. Entirely on its own, the robot arrives at the logical conclusion and, without any pre-programmed instructions, politely corrects itself. The robot, scientists say, BECOMES SELF-AWARE.
Indeed, none of these except the knowledge about the dumbing pill were pre-programmed, thus the experiment was “clean” — no trick or prank. The NAO robot was able to understand the question, recognize its own voice and come to the conclusion that it’s logically impossible it received the dumbing pill. (I catch myself trying to say he, instead of it when addressing genderless robots.)
However simplistic and crude, this AI test shows that artificial consciousness in its basic forms already exists, and gives us a glimpse into the behavior of a truly sentient robot of the future. Worthy of note also the fact that NAO aren’t even the most sophisticated robots, particularly the ones that took part in the experiment.
Thus beware and be warned: robots are really coming. Although Dr. Bringsjord claims that the human mind will never (like never-ever) be surpassed by that of a machine. Do I wish to live long enough to see him being proven wrong?