Researchers at the University of Tokyo have created a controversial robot by inserting lab-grown neurons from live rat cells to train artificial neural networks to acquire thinking skills and perform complex tasks, according to the Daily Mail.
This is the first time lab-grown cells from living cells have been used to train artificial intelligence machines.
The experiment was detailed in a paper called “Computation of physical reservoirs with FORCE learning in a live neural culture,” published in Applied Physics Letters, an article publication site of the American Institute of Physics.
The authors of the experiment said, “We developed a closed-loop system to generate a coherent signal from a spontaneously active living neuronal culture and embodied the culture with a mobile vehicle robot.”
The small wheeled robot was connected to a system of lab-grown nerve cells, which were stimulated to guide the robot to accomplish a specific task. In the experiment, the little machine had to reach a black circular box.
Whenever the robot missed the target, either by going to the wrong place, colliding with obstacles, or aiming its robotic vision away from the object it was supposed to reach, it was disturbed by electrical impulses. Thus the machine was guided.
During the test, the robot also received homeostatic signals indicating that it was progressing towards the target.
Thus, the robot received either homeostatic or disturbance signals (which caused it to wobble and recalibrate) depending on whether it was approaching or moving away from the goal it was supposed to reach.
Hirokazu Takahashi, the author of the study and associate professor of mechanoinformatics at the University of Tokyo, said:
“A brain of [an] elementary school kid is unable to solve mathematical problems in a college admission exam, possibly because the dynamics of the brain or their “physical reservoir computer” is not rich enough.”
The Japanese researchers believe that physical reservoir computing will contribute to a more detailed understanding of how the brain works, thus developing a neuromorphic computer, which could mimic the neurobiological fabric of a human’s nervous system.
But this kind of breakthrough in technology, called “physical reservoir computing,” could also be the beginning of the controversial creation of artificial intelligence machines that think like us, even displaying higher intelligence capabilities.
How far would science be able to go? It is the responsibility of societies to maintain a moral horizon to address ethical concerns related to some scientific experiments that affect human nature.