AI for All? Inclusive Learning Environments With and Through AI in Early Elementary Education

Abstract The aim of this paper is to identify the learning-promoting potential and key success factors of Artificial Intelligence (AI) for and in inclusive early elementary education. To this end, an intervention was designed and continuously refined over the course of three rounds (spanning 18 months) with the aim of promoting AI literacy among young children. Methodologically grounded in Design-Based Research (DBR), the intervention was implemented at an elementary school in Chemnitz/Germany with 41 children and was continuously refined. The intervention consists of three modules (how AI works; applications of AI; reflections on AI) and is outlined as an example in this paper. As part of the intervention’s development, the AI-related conceptions of elementary school children were assessed through group discussions and individual ebooks throughout the learning process. The results highlight the specific learning potential that the created learning environment as they illustrate that the participating children developed a basic understanding of AI, view AI as a digital-technical system, and describe AI using overarching concepts. The childen’s conceptions of AI can be charaterized by their view of AI as a learning system and described in contrast to (intelligent) humans. Furthermore, AI is viewed on the one hand as a knowledge-dispensing system, while on the other hand it is described as a flawed system. The results are discussed in relation to existing concepts for inclusive digital learning environments and framework models for AI literacy, which need to be adapted for elementary school children. The resulting intervention and the empirical findings on the children’s conceptions will serve as the basis for developing a Concept Inventory on AI literacy adressing elementary school children in early education.