Main » Projects » Artificial intelligence

Artificial intelligence

viewed 2560
Artificial intelligence

The purpose of the project

Describe a model of humanoid artificial intelligence.

Project tasks

Build a social-game model of natural intelligence; Formalize the main structural elements of the system of intellectual activity; Develop a schematic diagram of intellectual activity relevant to the means of machine implementation.

Annotation to the project

The content of our individual world of ideas is determined by the process of socialization, and its stability is ensured by the codes of brain structures; According to K. Anokhin, “The mind has a granular structure and consists of cogs - elementary units of experience that encode the relationship of the organism with the environment. Stable bonds between cogs are called loks. Connected loks (and cogs) form a network called cognitome (which is the bearer of all subjective experience of the organism)"; Empirical situations are encoded in experience, which are related to games in childhood and later associated with “social games”, to be later encoded as a “playful” universe by the cognitome, where things are always functions (derivatives) of the game; Since the game cannot exist “on its own”, the individual and intersubjective worlds built on top of them, as well as the way they exist (“being”) are “subject-sized”: the brain (and the “memory” of AI) encodes not “external things”, but ways of interacting with the "elements" of the outside world; The world for the intellect is a set of "game situations"; A person cannot find something in the “world” that is not connected with the semantic universe in which he was formed, since the “identification” of external objects and situations involves the actualization of the scenario of a social game known to him (out of the many social games existing in culture); Therefore, ideas are determined not by “canonical” (unchanging, objective) external “things”, but by game scenarios, and the question of the “appearance” of things is largely related to the “objectivity” (legitimacy) of the game. The world (and all its elements) is the other side of the "cognitome"; Therefore, AI should be implemented not as a self-learning neural network on billions of examples, but as a self-actualizing system of "social games".