This article examines the evolving discourse of creativity in the context of generative artificial intelligence (genAI), proposing the term MASS—Massive Abstract Statistical Systems—to critically reframe the role of genAI in creative practice. Drawing from literature across art education and computer science, the authors trace how creativity is conceptualized differently within these fields: as an embodied, culturally situated process in art education, and as a quantifiable, solvable question in AI research. Through comparative literature analysis of publications from arXiv, Studies in Art Education, and International Journal of Education Through Art, the study highlights tensions between human-centric, process-oriented notions of creativity and algorithmic, output-driven models characteristic of genAI.

The authors argue that the rise of genAI reconfigures creative agency, moving from intuitive and situated practices to data-driven processes shaped by prompting, statistical modeling, and computational abstraction. They critique popular narratives that frame genAI as inherently creative, asserting that such claims obscure its material and ecological costs, ethical concerns, and lack of intentionality or authenticity. MASS is introduced as a conceptual tool to shift discourse away from misleading anthropomorphisms of AI, emphasizing the system’s statistical logic and infrastructural implications.

The article concludes by calling for post-AI perspectives in art education—ones that critically engage with AI’s epistemologies, promote AI literacy, and foreground ethical and ecological considerations. Through this lens, creativity is not only redefined but re-worlded, opening new pedagogical and theoretical pathways for understanding the intersection of art, technology, and society in a computationally saturated era.

This study used a comparative literature analysis to examine how “creativity” is defined and operationalized across two intersecting knowledge communities shaped by generative AI: computer science research developing AI systems for creative applications and art education scholarship treating creativity as pedagogical, sociocultural, and embodied practice; the aim was not to measure creativity directly but to map the discourse of creativity—its assumptions, emphases, and omissions—and to identify tensions that become pronounced in a post-AI context. The authors constructed a corpus from three main sources: arXiv (as a high-volume, open-access repository representing AI development discourse), and two established art education journals—Studies in Art Education and the International Journal of Education Through Art (IJETA)—to represent disciplinary discussions of creativity in arts learning and teaching; they also used Google Ngram searches as contextual evidence of changing attention to “creativity” and the pairing “artificial intelligence + creativity” over time. For arXiv, the search strategy targeted the keywords “creative artificial intelligence” and “art” within computer science, using “art” to refine relevance and screening the initial returns to focus on creative-domain work; this process yielded 55 arXiv articles announced in 2024 that met the stated criteria. For the art education journals, the authors searched titles, keywords, and abstracts for “creativity,” then excluded articles centered on the “creative industry” or those not substantively engaging creativity as a concept; after exclusions, the art education sample comprised 52 articles (34 from IJETA and 17 from Studies in Art Education, as reported). The analysis proceeded qualitatively, comparing how texts define creativity, what frameworks they invoke, and whether creativity is treated primarily as process, product, pedagogy, a measurement problem, or situated sociocultural practice; findings were then synthesized into thematic discussions (including GenAI and creative practice, differentiating creative applications of AI, and sociocultural implications of GenAI) to support cross-field “boundary work” that surfaces persistent divergences—such as quantification versus embodiment and output evaluation versus situated process—informing the paper’s broader conceptual claims.

The study finds a pronounced disconnect between how creativity is framed in AI research versus art education scholarship: across the reviewed arXiv corpus, creativity is commonly treated as a measurable, optimizable problem—something that can be improved through the right procedures, models, and evaluation metrics—whereas art education literature understands creativity as a relational, culturally situated, and embodied assemblage of human and non-human activity that cannot be reduced to output scoring alone. The authors also observe that creativity-focused discourse has surged in AI development communities (e.g., arXiv outputs and related trend indicators), while art education research on creativity appears comparatively outpaced—an imbalance that matters because computational definitions of creativity risk becoming dominant in public and institutional understandings of what it means to create. As a critical outcome, the article proposes reframing generative AI as MASS (Massive Abstract Statistical Systems) to counter misleading anthropomorphisms of “intelligence” and “creativity,” emphasizing instead the statistical-logical nature of these systems and their material and infrastructural entanglements. In terms of key findings for post-AI art education, the review highlights recurring tension points requiring “tactical intervention and critical scholarship,” particularly debates over authorship and agency in co-creation, intellectual property and claims of “labor theft” tied to training data, algorithmic bias and cultural misrepresentation, and the comparatively weak attention to ecological costs (energy, raw materials, and sustainability concerns) within AI-facing creativity discourse. The authors conclude by calling for post-AI perspectives in art education that build AI literacy while foregrounding ethical, cultural, and ecological considerations, asking how creative practice and arts learning are being reshaped now that LLM-driven tools are embedded in everyday creative workflows—and what those practices should become.

Allred, A. M., & Aragon, C. (2023). Art in the machine: Value misalignment and AI “art”. In International Conference on Cooperative Design, Visualization and Engineering (pp. 31–42). Springer. Anderson, R. C. (2018). Creative engagement: Embodied metaphor, the affective brain, and meaningful learning. Mind, Brain, and Education, 12(2), 72–81. Anderson, T., & Guyas, A. S. (2012). Earth education, interbeing, and deep ecology. Studies in Art Education, 53(3), 223–245. https://doi.org/10.1080/00393541.2012.11518865 Banaji, S., Burn, A., & Buckingham, D. (2010). The rhetorics of creativity: A literature review (2nd ed.). Creativity, Culture and Education. Craft, A. (2001). An analysis of research and literature on creativity in education. Qualifications and Curriculum Authority. Crawford, K. (2021). Atlas of AI: Power, politics, and the planetary costs of artificial intelligence. Yale University Press. Demir, C. (2005). Enhancing creativity in art education through brainstorming. International Journal of Education through Art, 1(2), 153–160. https://doi.org/10.1386/etar.1.2.153/3 Garoian, C., & Gaudelius, Y. M. (2001). Cyborg pedagogy: Performing resistance in the digital age. Studies in Art Education, 42(4), 333–347. https://doi.org/10.1080/00393541.2001.11651708 Gay, C. H., Figueroa-Sarriera, H. J., Mentor, S. (1995). The cyborg handbook. Routledge. Goetze, T. (2024). AI art is theft: Labour, extraction, and exploitation: Or, on the dangers of stochastic Pollocks. In Proceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparency (pp. 186–196). ACM. https://doi.org/10.1145/3630106.3658898 Lewis, T. (2024). The creative act in the studio: A plea for rethinking potentiality in art education. International Journal of Education Through Art. 20. 337-350.10.1386/eta_00172_1. Runco, M. A. (2023). AI can only produce artificial creativity. Journal of Creativity, 33(3), 100063. https://www.sciencedirect.com/science/article/pii/S2713374523000225 Williamson, B. (2024, November 8). Critical keywords of AI in education. https://codeactsineducation.wordpress.com/2024/11/08/critical-keywords-of-ai-in-educ ation/ Williamson, B., & Eynon, R. (2020). Historical threads, missing links, and future directions in AI in education. Learning, Media and Technology, 45(3), 223–235. https://doi.org/10.1080/17439884.2020.1798995