Cosentino, G., Anton, J., Gelsomini, M., Sharma, K., Giannakos, M., & Abrahamson, D. (2025). From teachers to AI: How human-student interactions inform AI feedback design for embodied math education.

In “Augmented educators and AI: Shaping the future of human-AI collaboration in learning.” Workshop presented at CHI 2025—the annual meeting of the Association of Computing Machinery (ACM) special interest group Computer–Human Interaction. ACM.

ABSTRACT: The integration of Artificial Intelligence (AI) in education is transforming teaching and learning dynamics. AI-powered tools have the potential to enhance teaching by supporting and augmenting instructional practices. Human-AI collaboration in classrooms can significantly strengthen personalized learning, improve student engagement, and optimize formative feedback. However, for AI to effectively support teachers and students, it must exhibit human-like qualities such as empathy, adaptability, and contextual awareness. This study examines how children interact with AI in an embodied learning environment for mathematics education. Using a body-scale digital number line, students enacted integer arithmetic procedures through physical movement. A between-group design was implemented: one group received feedback from a human teacher, while the other received AI-generated feedback. A mixed-method approach combined multimodal data (eye tracking) with qualitative analysis of student interactions to feedback. Findings highlight the potential of incorporating pedagogical strategies into AI design and underscore the complementary role of AI and human teachers, highlighting key design considerations for AI systems that aim to balance automation with human-like communication in education.