Metasystem Transitions in Education: Human-AI Interactions as Catalysts for Computational Thinking

BRAIN. Broad Research In Artificial Intelligence And Neuroscience (Romania). 2025. Vol. 16. No. 4. P. 72–85.

Patarakin Y., Korchak A., Burov V., Costley J.

The paper explores how human–AI interactions, particularly with generative artificial intelligence (Gen-AI), serve as catalysts for developing computational thinking (CT) as a creative meta-competence. Grounded in a theoretical framework that synthesises Vygotsky’s mediated learning, Lotman’s Semiosphere, and Turchin’s Metasystem Transition theory, the study conceptualises education as a process of collaborative meaning-making within human–machine teams. Two case studies conducted in Russian middle schools provide empirical grounding: one involving scaffolded Gen-AI use in project-based learning, and the other involving unstructured Gen-AI use during programming tasks. Results indicate that structured, reflective engagement with Gen-AI enhances students’ CT skills and supports self-regulation, while not scaffolded use yields mixed outcomes. The findings suggest that Gen-AI can function as a semiotic and cognitive scaffold when meaningfully integrated into educational activities. The study contributes to the understanding of computational thinking as a form of cultural and cognitive mediation, and positions Gen-AI as a driver of metasystem transitions in education.

Keywords: computational thinking, generative artificial intelligence, mediated learning, metasystem transition