The Impact of Large Language Models (LLM) on Conceptual Approaches to Creativity

- Тадеєв, Юрій Петрович (orcid.org/0000-0002-1148-4798) (2025) The Impact of Large Language Models (LLM) on Conceptual Approaches to Creativity Pedagogical Innovations: Ideas, Realities, Perspectives, 2 (35). pp. 133-137. ISSN 2413-4139

Text ПІ(2)35-2025_13-01-133-137.pdf Download (365kB)

Abstract

This article provides a comprehensive theoretical analysis of how large language models transform contemporary approaches to defining creativity. The study draws on classical psychological, philosophical, and sociocultural theories, particularly concepts of divergent and convergent thinking and the distinction between P-creativity (psychological) and H-creativity (historical). The research employs conceptual analysis, integrating insights from psychology of creativity, philosophy of science, pedagogy, and AI studies to examine methodological shifts in understanding creativity as both psychological and sociocultural phenomenon. The analysis reveals ambivalent effects of LLM integration into creative, educational, and scientific practices. While these models enhance fluency and short-term productivity, they introduce significant risks: homogenization of thinking through reproduction of dominant patterns, reduction of autonomous creative capacity, and transformation of authorship in human-machine collaboration. LLMs effectively support P-creativity by expanding usersʼ cognitive space, yet their H-creativity capacity remains limited as they operate within existing linguistic patterns. A critical finding concerns the validity crisis in traditional psychometric creativity assessment. Conventional instruments, including Torrance Tests, require methodological revision to account for algorithmic support. Quantitative increases in idea generation do not guarantee enhanced originality, as LLM outputs favor statistical probability over genuine innovation. The paper outlines future research directions in developing process-oriented assessment models that account for human-AI interaction specifics in educational and scientific contexts.

Item Type:	Article
Keywords:	creativity, large language models, divergent thinking, convergent thinking, artificial intelligence.
Subjects:	Science and knowledge. Organization. Computer science. Information. Documentation. Librarianship. Institutions. Publications > 1 Philosophy. Psychology Science and knowledge. Organization. Computer science. Information. Documentation. Librarianship. Institutions. Publications > 3 Social Sciences > 37 Education > 37.01/.09 Special auxiliary table for theory, principles, methods and organization of education > 37.01 Fundamentals of education. Theory. Policy etc.
Divisions:	Institute of the gifted child > International Collaboration and Fostering Giftedness Depatment
Depositing User:	н.с. Олена Василівна Онопченко
Date Deposited:	28 Jan 2026 18:41
Last Modified:	28 Jan 2026 18:41
URI:	https://lib.iitta.gov.ua/id/eprint/748299

Downloads

Actions (login required)

View Item