Can generative ai be a partner in student decision-making in computational physics?

As generative AI (genAI) tools become increasingly common in classrooms, questions arise about how they might influence the ways students think, create, and make decisions. In this work, we examine how students in a computational physics lab use genAI in their coursework, focusing on how their interactions with the tool reveal shifts in agency and expressions of creativity. Drawing on qualitative field notes and iterative coding, we present a developing codebook that characterizes patterns of genAI use and student decision-making. We observed that students decision-making processes take shape through how they choose to prompt genAI and how they interpret and respond to its generated outputs. This study contributes to the ongoing conversations in PER and in physics course environments about how emerging technologies influence student learning, agency, and creative practice.