Exploring Pathways to Creativity in Early Childhood through AI-Enhanced Musical Play

• In Event: Computational Playgrounds: Foundations and Futures of AI, Play, and Pedagogy in Early Childhood

Abstract

Objectives
This study explores how artificial intelligence (AI)–enhanced musical play can support creativity, cultural awareness, and historical inquiry in early childhood classrooms. Specifically, we investigate how a digital remix platform—Citizen DJ—can be used by preschool children and their educators to remix historical audio artifacts and engage in multimodal music-making as a form of identity expression, inquiry, and cultural connection.
Theoretical Framework
The study is guided by constructionist theory (Papert, 1980), which emphasizes learning through building and designing personally meaningful artifacts. We also draw on culturally responsive pedagogy (Broughton, 2017), distributed creativity (Sawyer & DeZutter, 2009), and the affordances of play-based, multimodal learning in early childhood (Barrett et al., 2022; Nieuwmeijer et al., 2021). These perspectives frame how children’s remixing activities function as acts of cultural and creative agency within sociocultural contexts.
Methods
A qualitative case study was conducted over four weeks in a preschool classroom with 15 four-year-old children, one lead teacher, and two teaching assistants. The implementation followed three progressive phases: (1) Introduction and Familiarization, where children explored how to remix sounds using Citizen DJ; (2) Guided Play and Exploration, with scaffolded small-group sessions emphasizing tempo, volume, and sound combinations; and (3) Independent Creation and Reflection, where children composed original remixes and discussed their creative decisions with peers. Educator facilitation emphasized inquiry and responsive engagement rather than performance.
Data Sources
Data collection included video recordings, field notes, educators’ reflective journals, and semi-structured interviews with children. Children’s digital artifacts—original sound remixes using Citizen DJ—were also collected. The platform, which uses AI to organize and classify a large archive of public domain audio (e.g., spoken word, rhythms, environmental sounds, historical recordings) based on thematic and qualitative attributes, enabled children to search, select, and remix audio content through an intuitive interface.
Results
Findings indicated that children demonstrated originality and divergent thinking in their compositions. Metrics such as marginal distinct idea count and non-redundant idea count were used to evaluate novelty in remixing. Network analysis of sound selection showed that children moved fluidly across thematic sound categories—particularly favoring spoken word and animal sounds—which often initiated questions about the origins of the audio content, prompting cultural and historical dialogue. A Cultural Engagement Heatmap highlighted frequent engagement with culturally meaningful content, such as folk music, archival speech, and nursery rhymes. Educators noted increased confidence in incorporating music-based activities and reported that Citizen DJ expanded their pedagogical toolkit to include technology-mediated cultural inquiry.
Scholarly Significance
This study offers novel contributions to early childhood educational research by demonstrating how AI and machine learning can facilitate playful, developmentally appropriate engagement with historical sound archives. Citizen DJ serves as a model for democratizing access to cultural heritage and scaffolding young children’s creative agency. The platform supports both child-directed learning and educator-facilitated exploration, fostering sound-based inquiry that nurtures creativity, identity development, and historical awareness. These findings illuminate the pedagogical potential of AI-driven tools in transforming music learning and expanding cultural literacy in the early years.

Authors

• Ilene R. Berson, University of South Florida

• Michael J. Berson, University of South Florida