Using GenAI in complex tasks: catalysing self-evaluation, learning through feedback, and iterative collaboration in chemical engineering design

Through projects created by MSc/MEng Chemical Engineering students, we explore collaboration between learners and genAI tools in activities of differing degrees of complexity, from a decision analysis assessment to an open-ended process design. Interacting with the AI tool (ChatGPT) leads students to a deeper critical evaluation and improvement of their own understanding, by provoking their own feedback on the AI’s performance. Particularly in the more complex multistage design task this leads to an iterative use of the AI tool to refine the design, combining the student’s advancing understanding of the task with increasingly focussed, useful ‘help’ from the AI.