As researchers increasingly work on artificial intelligence (AI), they discover newer facts every other day. We have seen the technology used to assist the military, control traffic systems, beat humans in video games, and even control a ship. Now, a recent study has found that a neural system needs to have a few hours of “sleep” to function efficiently.
The discovery was made by a team of researchers from the Los Alamos Laboratory located in the United States. These researchers, while working on an AI project, found that the neural system became enormously more effective after artificially induced sleep.
Thus, while working on the neural system for prolonged hours, the researchers first found that the system is becoming unstable. He tried to perform his task abnormally as he tried to classify the objects using their dictionary definitions without having any example to compare them with.
However, after the researchers exposed the neural system to an artificial analog of sleep, it returned to normal conditions, performing its task efficiently again. Presently, this artificial analog of sleep is, according to researchers, similar to a human’s nighttime sleep.
Garret Kenyon, a computer researcher working on the project, said: “The question of how to prevent learning systems from becoming unstable only arises when trying to use biologically realistic, tipped neuromorphic processors or trying to understand biology itself. This wide majority of machine learning, deep learning, and artificial intelligence researchers nevermore find this problem because in the highly artificial systems they study they have the luxury of performing global mathematical processes that have the impact of controlling the overall dynamic gain of a particular system. “