This paper investigates the impact of AI-assisted programming (GitHub Copilot) versus human pair programming on novice programmers' performance, learning, workload, and emotional experiences. Findings indicate that while AI improves performance and reduces workload, human pairing fosters more positive emotional engagement and potentially better long-term learning retention, suggesting a need to re-evaluate traditional pair programming in education.

This study empirically investigates how school-level Artificial Intelligence (AI) readiness influences student AI literacy in vocational education through the mediating role of aggregated teacher capabilities. Using a large national dataset, it establishes that institutional preparedness, particularly when translated into teacher competence, significantly impacts student learning outcomes across diverse regional contexts.