Epistemological transformations of mentality and semantic coworking with large language models

Abstract

The article carries out a philosophical, anthropological and epistemological analysis of the transformations of human mentality in the conditions of the spread of large language models and generative artificial intelligence. The influence of LLM on the structure of cognition, the formation of meanings, cognitive practices and creative experience of modern man is studied. The thesis is substantiated that generative artificial intelligence is not only a technological tool, but a factor in a profound change in the way of human existence in digital culture, a marker of the emergence of a new environment as a primarily cognitive space. Special attention is paid to the analysis of the transformation of the classical model of the subject of knowledge, which gradually moves from autonomous thinking to algorithmically mediated forms of cognitive activity. It is determined that large language models change the nature of access to knowledge, strengthen tendencies towards fragmentary perception of information and form a new type of epistemic addiction. The paper considers the positive and negative aspects of the use of generative artificial intelligence. Positive trends include the democratization of access to knowledge, the development of personalized learning, the acceleration of interdisciplinary synthesis and the expansion of opportunities for creative interaction between humans and digital systems. The main risks include the standardization of thinking, the superficiality of cognitive practices, the devaluation of intellectual effort, the crisis of authorship and the strengthening of information manipulation. The phenomenon of algorithmic mediation of reality is analyzed, within which a significant part of human social experience is formed through texts and semantic constructions created by artificial intelligence. A special emphasis is placed on the problem of creativity and the change in the psychology of the creative process in the conditions of digital culture. LLMs do not possess existential experience, corporeality or intentionality, but they are able to act as a tool for cooperative creativity and stimulate new forms of creative interaction. The article proposes several possible models for the development of human interaction with generative artificial intelligence: adaptive-humanistic, technocratic, symbiotic and crisis. The concept of semantic coworking is also substantiated and the need to form a humanistic strategy for the coexistence of humans and algorithmic systems is concluded, at the center of which should remain critical thinking, cognitive autonomy, and the value of human experience.