The consciousness or ability that humans have to recognize our own existence and to imagine ourselves, is one of the qualities that separates us from the robots we know, from those we see in science fiction movies.
However, there is a group of engineers from Columbia University, who claim to be much closer than before the creation of an artificial consciousness, having been able to develop a robot capable of imagining itself, and to carry out complex tasks thanks to its own internal simulations, and not to predefined algorithms with instructions and sequences written by its creators, as is usually the case with most robots.
“Until now, robots have operated following a model explicitly designed by a human. But if we want robots to become independent, to quickly adapt to unforeseen scenarios by their creators, it is essential that they learn to simulate themselves “, commented Hod Lipson, professor of mechanical engineering and director of the Creative Machines laboratory at the university. from Columbia.
The birth of artificial consciousness
To test their theory, the engineers used a robotic arm, which at the beginning of the experiment knew nothing about itself, neither its shape, nor its capabilities. After a period of “meditation” of about 35 hours, where the robot in a rudimentary way was putting together a model of its own shape and possible movements, managed to put together a model of himself quite accurate with only four centimeters of difference compared to its actual measurements.
Once the robot could imagine itself, it was given a task in which it had to repeatedly move an object from one side to the other. Thanks to its own internal simulations, the robot could imagine its own movements before performing them, perfecting his real movements with each attempt until he finally managed to perform the task perfectly and without the need for instructions from its creators.
To complicate matters, the researchers even changed a part of the robot’s body to make it misshapen from its original shape, and so on. simulate for example a fault in one of its parts, which resulted in a very rapid adaptation of the internal model of the robot, affecting its performance very little. Something that would have been practically impossible with a traditional robot that only follows instructions.
Although its creators are aware that this model of artificial consciousness it’s quite rudimentary Compared to what we humans do, they warn that it opens the doors to a new era in robotics where they even warn that new ethical considerations are going to be born in between: “Self-awareness will lead to more resilient and adaptive systems, but it also involves some loss of control. It’s a powerful technology, but it must be handled with care“.