With the rapid advancement and broad availability of large language models such as ChatGPT, students and professors alike have been thrust into a gray zone of decision-making: when is it okay to outsource reading and writing to A.I., and what are the implications for learning, individually and collectively? In a graduate introductory level plant pathology class, we implemented an assignment set that guided students through both ethical and quality concerns with A.I., addressing questions such as: How does use of A.I. impact scientific research in our discipline? How do we measure quality, and can A.I. tools create quality work or evaluate quality in submitted work? What do we gain and lose by having A.I. read and summarize for us? What is my own process for reading and understanding research articles, and what specific roles might A.I. play in helping me locate and assess relevant information? How do I mitigate the risks of bias, omission, or hallucination inherent in the LLM platforms? Students progressed through one assignment every three weeks, allowing time for experimentation and reflection with each assigned activity and question set. In reflections and discussions, students expressed a range of initial attitudes towards A.I. use, and demonstrated growth in their evaluations of the risks and benefits of LLMs. They were able to connect their own experiences to collective impacts (e.g., the production-progress paradox), demonstrate the pitfalls of A.I. driven metrics, articulate their own rationales and parameters for LLM use, and begin to address the ultimate question: why should someone hire you, instead of ChatGPT?