Reflections on AI in mechanical engineering education

Abstract

Most current applications of AI in mechanical engineering emphasize ChatGPT-like tools for summarization or explanation, often overlooking AI's potential to guide both learning and research through structured navigation of complex problem spaces. This paper presents personal reflections on how AI can transform engineering pedagogy by fostering deeper learning, guiding students to actively construct understanding, and progressively supporting their development from guided exploration to independent reasoning. This shifts the emphasis from procedural mechanics (removing steps, sequencing tasks) to the cognitive journey of the learner, highlighting engagement, understanding, and independent thinking. By leveraging principles from Cognitive Load Theory—managing intrinsic, extraneous, and germane load—AI can dynamically tailor guidance to learners’ evolving understanding. While this requires additional effort from the instructor, the payoff is increased engagement. The same guiding principles extend naturally to research and computational workflows. Rather than replacing research or numerical solvers, AI can operate within established frameworks—such as finite element methods and rigid multi-body dynamics—to identify physically meaningful starting points, navigate families of related solutions, and accelerate analysis while preserving theoretical integrity. In both educational and research contexts, AI functions not as a black-box solver but as a navigator of structured solution spaces. These reflections are intended as conceptual insights and practical proposals rather than a comprehensive literature review. In an era of rapid technological change, this approach—using AI as a guide for both learning and computation—offers a coherent framework for deeper engagement, future research, and responsible integration into engineering education and practice, while formal assessment remains in abeyance.