Digital technology such as No Code Gen AI using Large Language Models (e.g., ChatGPT) has high potential in shaping our higher education. However, there are prevalent concerns that ChatGPT may encourage students to take shortcuts and become lazy, resulting in surface or no learning. Addressing this concern, we have developed a “Tool-to-Teammate” instructional strategy for integrating ChatGPT in teaching and learning. Students enrolled in a graduate program, “Urban Science Policy and Planning,” at a university in Singapore on Creating the Frontiers of the No Code Smart City were tasked to design a future AI classroom as part of their classroom activities using ChatGPT. The course involved 17 graduate students divided into three groups. Two groups were provided with pre-built Custom GPTs that focused on either Observation or Interview Skills. In the first phase, the two groups familiarized themselves with their respective Custom GPT as a "Tool”. Subsequently, these groups collaboratively enhanced their Custom GPTs, shifting their interaction with the AI as a "Teammate." The third group, serving as a control, was tasked with creating their own Custom GPT from scratch, subsequently employing it as both Tool and Teammate in designing their future classroom. Finally, students presented their co-developed custom GPT to discuss as a class and learn. We used Observations, student reflections, and post-class semi-structured interviews to understand student experience on using ChatGPT as a Tool and Teammate. Preliminary results indicate that students found ChatGPT to be engaging. They reflected that it allowed them to be creative while enhancing their problem-solving skills to deliver solutions. We discuss the possibilities and implications of the “Tool-to-Teammate” instructional strategy for using ChatGPT in Teaching and Learning.

Keywords: Tool-to-Teammate Instructional Method, ChatGPT, Teaching and Learning, Higher Education.