This paper is devoted to the analysis of the culture of dialogue in the digital age and the influence of language models such as ChatGPT on the educational process. The essence of the ancient dialogue is described as the goal and method of a joint search for truth. The models of communication “human-human” and “human-artificial agent” are compared. It is stated that the main purpose of ChatGPT is to generate plausible-sounding answers, and not to seek objective truth or engage in genuine dialectical research. The example of a conversation with a chatbot shows the limitations in generating quasi-philosophical texts. Nevertheless, the pedagogical potential of ChatGPT as a tool for the formation of dialectical and critical thinking is indicated. Socratic methods for teaching people and machines (dialectic, maieutics, definition, elenchus, etc.) are considered. Two modes of operation of ChatGPT are described – the “captive” mode (Plato's cave) and the “mechanical oracle” mode (artificial Socrates).
Artificial Intelligence (AI); Large Language Models (LLM); ChatGPT; philosophy; education; dialogue; dialectic; Socratic Method.