Textbooks to Chatbots: The Artificial Intelligence Shift in Learning Heat Treating Failures

Heat treatment is a foundational topic in materials science, and requires a strong grasp of atomic structures, phase transformations, and failure mechanisms. Traditionally, students have relied on textbooks, lectures, past exams, and collaboration to build this understanding -- a slow but rigorous process. However, the rapid adoption of AI has fundamentally changed how students learn these concepts. As a member of the UConn MSE Class of 2025, I have had the unique experience of learning materials science both before and after AI became widely accessible. While AI serves as a valuable tool for reinforcing concepts and reducing time spent on grunt work, it has also become a crutch for many students, weakening critical thinking skills in engineering.

This presentation explores the double-edged nature of AI in learning heat treating failures. AI can assist in brainstorming failure mechanisms, such as testing thought paths through a hypothetical failure analysis. However, the effectiveness of AI in a given situation depends on the user’s ability to critically assess its output. As a teaching assistant, I have seen a growing difficulty among students to distinguish accurate information from confident-sounding AI-generated redundancy or errors. The challenge lies in developing the metallurgical intuition to quickly recall concepts like phase diagrams, CCT/TTT curves, and failure case studies without excessive reliance on external tools.

AI has a place in failure analysis and engineering education, but its role should be carefully integrated. Used appropriately, it can streamline workflow and aid in project development, but overuse risks impacting students' foundational knowledge necessary for sound engineering decisions. This talk invites discussion on the benefits and pitfalls of AI in heat treating education and strategies for its responsible use in developing the next generation of failure analysts.