A researcher at Rensselaer Polytechnic Institute in New York has administered a “wise man” puzzle to three humanoid robots in an effort to establish an element of self-awareness in the machines.

In the test, the three robots are told that two of them have been given a “dumbing pill” that stops them from talking. In reality, researchers have silenced them with the push of a button, but none of the robots knows which one is still able to speak.

Showing that robots can tackle logical puzzles requiring an element of self-awareness is a step toward building machines that understand their place in the world.When they are unable to solve the problem, the robots all attempt to say “I don’t know.” But only one of them makes any noise. Hearing its own robotic voice, it understands that it cannot have been silenced. “Sorry, I know now! I was able to prove that I was not given a dumbing pill,” it says. The robot then writes a mathematical proof and saves it to its memory to prove it has understood.

Researchers says that showing that robots – in this case, off-the-shelf Nao models – can tackle logical puzzles requiring an element of self-awareness is an important step toward building machines that understand their place in the world.

Selmer Bringsjord of Rensselaer Polytechnic Institute, ran the test and says that by passing many tests of this kind robots will build up a resume of abilities that start to become useful. The work, which was presented at the RO-MAN conference in Kobe, Japan.