A team of scientists from the Rensselaer Polytechnic Institute AI and Reasoning Lab in New York gave an adapted version of the puzzle “The King’s Wise Men” to three robots, and one of the robots was able to solve the puzzle while showing signs self-awareness.
Selmer Bringsjord, head of the department of cognitive science at the institute, will present the research in Kobe, Japan later this year at the IEEE symposium on robot and human interactive communication, the International Business Times reported.
Three humanoid Nao robots from the French robotics company Aldebaran were used in the self-awareness test. All three robots were made aware of a “dumbing pill” — a button on top of their head that would render them silent, according to SlashGear. All of the robots then had their heads tapped on by the tester, although only two were actually “given the pill.”
Then the robots were each asked which pill they received. Two of the robots didn’t speak, due to the dumbing pill. They remained silent, their human-like eyes seemingly watching each other, and the tester, to see if they were correct.
The third robot, after slowly getting up as if to make an announcement, simply replied, “I don’t know.”
Then after a pause, it added, “Sorry, I know now. I was able to prove that I was not given the dumbing pill,” while moving its arms about in dramatic gestures.
Sorry, I know now. I was able to prove that I was not given the dumbing pill.
Surprisingly, none of these answers were pre-programmed into the robots. The only information they were programmed with was the knowledge that the dumbing pills would make them unable to speak if it was placed on their heads.
In other artificial intelligence news, a cockroach robot in California has learned how to navigate tight spaces and maneuver around obstacles.
