Search for a command to run...
This paper presents a comparative analysis of the causal factors underlying information processing in the neuronal networks of the human brain and in the electrical networks of contemporary artificial intelligence (AI) systems, by physical processing. It is shown that both in the first and in the second of the listed cases the factor of a person's ability to identify the subjective value of information acts as an information-processing operator, that is, as a factor of controlling, determining (orchestrating) the information process. It was found that in the case of AI systems, it should also be considered that the reason and driver of the information process in them is, despite their entirely physical nature, the subjective value of those information processes, which is initiated by a person within the framework of biological or social values. Next, the general information -processing mechanism in neuronal networks of the human brain is described, which in this paper (using the well-known terms of Gazzaniga and Tononi) is called "Bottom-Up and Top-Down mechanism of integration of past experience (information) through subjectively active mental phenomena." Here it is argued that a General Theory of Brain Information Activity cannot be created without considering the functional role of mental phenomena in the brain. The point of view is advocated that, by embodying integrated experience through mental phenomena into the structure of voluntary motor acts, the human subject (agent) can change the surrounding physical world from states of an "undesired present" to states of a "desired future". And this, in turn, testifies to the deep functional and causal connection of physical processes and the psyche, pointing to the modelling (through the phenomenon of information) nature of the psychic. Based on this statement and considering the totally physical nature of modern AI systems, it is argued that the latter remain governed by the subjective values of a person (developer, programmer, and user). The very functional architecture of modern AI remains as a system of choosing one of the most probabilistically correct outputs ("answers", "decisions") on a continuum of combinatory possibilities of a specific technical device, but, in any case, through the subjective values of a person.