The next robot you come across could be one that thinks like you
Research reveals that physical reservoir computing could make a robot think like human beings.
Science has grown in leaps and bounds over the past century and has come to a point where any dream can be transformed into reality. It was actually wonder when we were first introduced to robots. Scientists are now on a quest to make robots think like humans!
But can they actually make it happen? Research is proving that they could do that, indeed. Science has hinted that physical reservoir computing when pushed further could make artificial intelligence machines think like human beings. Physical reservoir computing is in real terms a technology that makes sense of brain signals.
Making a robot think
Researchers from the University of Tokyo have established that a robot could be trained to navigate through a maze by electrically stimulating a culture of brain nerve cells connected to the machine, a Techxplore report quoting a paper in the Applied Physics Letters, has said.
As per the researchers, the nerve cells, also known as neurons, were grown from living cells. They also were deployed as a physical reservoir for the computer to construct coherent signals. These neurons acted as the steering mechanism of the robots.
A maze task was assigned and, at times, when the robot moved off the mark, an electric impulse caused due to the disruption in neurons tried to correct its course. After a flurry of attempts, the robot was able to find its way through the maze.
Sending disturbance signals to set goals
This gave the researchers the confidence that such goal-directed behaviour could be cultivated. Also, it was found that such goal-directed behaviour can be ensured sans any additional learning by sending disturbance signals to an embodied system, the report added. The robot has entirely banked on electrical trial-and-error impulses.
The research had banked on the knowledge that a living system boasted intelligence through the mechanism of extracting a coherent output from a chaotic state.
This principle proved that intelligent task-solving abilities can be generated by deploying physical reservoir computers and that would help extract chaotic neuronal signals.
These signals could then cause homeostatic or disturbance signals, thereby aiding the computer to have a reservoir that understands how to solve the task.