Remember all those movies you’ve watched about that robot with artificial intelligence that wanted to destroy the world? Luckily, we don’t have to worry about that (yet), but artificial intelligence is still something that’s being worked on and it’s making progress. One of the things lacking in today’s artificial intelligence that would create a doomsday scenario is self-awareness. Self-awareness is the knowledge of one’s own presence, desires, and feelings. Rensselaer Polytechnic Institute professor Selmer Bringsjord has been testing AI with a few NAO robots and just demonstrated one of them becoming self-aware for a moment.
In the video below, the professor is using a variant of a logical puzzle commonly known as the “King’s Wise Men“. In the test, three robots have been made aware of something called a “dumbing pill” which is actually simply a button on the top of their heads which silences the robots. After that, Bringsjord taps each of their heads, but only presses the button (the pill) on two of them. After that, he asks them which pill they received.
After being asked that question, the robot who did not “get the pill” stands up and responds that he does not know. A couple seconds later, he says “Sorry, I know now. I was able to prove that I was not given the dumbing pill”.
What does that all mean? After hearing his own voice, the robot uses logic to discern that he was not given “the pill” and states as such. While this is no reason to panic, it does show that the robot was self-aware for a brief moment. While this is a crude form of self-awareness, it is the beginnings of something greater.
So, panicking yet?