The point of learning is for AI to learn to analyze and describe human behavior. Experts believe that the best means for this are literary texts, on the basis of which neural networks will be able to study human behavior.
It is based on the concept of the American psychologist Kurt Lewin, who proposed breaking stories or other works into fragments, including descriptions of the character’s social status, his psychophysiological state, circumstances and behavior, and then analyzing the connections between these variables.
As a result, the AI itself will learn to write something similar. In the meantime, technology writes primitive texts. According to scientists, because people themselves do not understand well the cause-and-effect relationships of their behavior.
Scientists hope that in the future the idea will interest programmers who will create a neural network that can analyze literary texts and help teach AI to better understand human behavior.