Scientists in Japan have taught a robot how to ‘think’ like a human using a Frankenstein-like technique!

Researchers at the University of Tokyo created a number of experiments for their tiny robot which ‘taught’ it how to find it’s way through a maze.

The robot works by using brain-like nerve cells that have been grown in a laboratory, and combining them with a tiny computer. This technique is called “physical reservoir computing”.

Brain neurons, also known as nerve cells, are responsible for sending information and messages in our brains.

How do you teach a robot to ‘think’?

To teach the robot to think, scientists created an experiment in which the robot had to find it’s way through a maze to reach a prize – a black round box of lights.

Throughout the robot’s journey, if it went in the wrong direction, or bumped into an object, the researchers would send an electrical signal through the robot’s neurons, to help it to learn more about it’s environment and put it back on track.

This is the first time intelligence has been ‘taught’ to a robot, and the researchers think it is a big step towards creating machines that will solve problems by thinking like humans.