Artificial Intelligence for Teaching and Learning Computational Physics: A State-of-the-Art Review of Trends, Challenges, and Future Directions

Abstract

Artificial intelligence (AI) has recently been recognized as playing an increasingly strategic role in computational physics education, notably with respect to the current strong focus on developing students’ ability to comprehend and apply computational modeling, data simulation, and problem solving within modern physics curriculums. This paper presents a state- of-the-art narrative review on the integration of AI in computational physics and related educational contexts, considering research from 2019 to date. The review discusses prominent AI technologies, pedagogical integration strategies and their documented effects on computational thinking, numerical modeling and student engagement. The synthesis also suggests that AI enhanced learning environments can support personalized feedback, adaptive assessment and flexible learning pathways in order to promote students‟ engagement and computational reasoning of physics learning. Meanwhile, the literature also shows ongoing challenges: ethical and equity issues; algorithmic bias; lack of instructor preparation for OAI pedagogies; and shortage of longitudinal empirical evidence in computational physics education. Through summarizing prevailing research trends and framing a set of important constraints, this review is intended to serve as a conceptual blueprint for an expanded research agenda towards the systematic, responsible, and pedagogically informed inclusion of AI in computational physics education