“Large Language Models in Education” Project Updates

Our UNIS colleagues share their ideas for integrating Large Language Models into their teaching.

We’ve been busy. Our first report is now out, and we have had two more workshops and two recent talks. In mid-November, we met with 15 UNIS staff to discuss the challenges and opportunities associated with the use of Large Language Models (LLMs) such as chatGPT. The day concluded with each participant presenting an implementation plan for how they would use these tools in their future teaching. They had some really inspired ideas! We also recently met with the group at UiB, following up on our initial workshop in August this year. We heard from these instructors about their recent experiences with LLMs in the classroom, and then discussed recent data—from the bioCEED national survey—on student perceptions of LLMs. After identifying several themes related to student concerns, we drafted a small report that describes these student concerns and offers some teacher responses to them. For example, many students were concerned about what constitutes acceptable use of LLMs (”I would have liked some training in how to use these tools without falling into the trap of cheating or incomplete work.”) To this, our group responded “Each instructor will have different rules and expectations. If these aren’t clear, ask! Each instructor is responsible for communicating the same set of expectations to the entire class.” This report will soon be finalized and shared through our project page and bioCEED. We presented on our findings, and these workshops, at the recent ISSOTL conference in Utrecht, and at the recent (8 December) studiekvalitetsseminar at UiB.