Using LLMs for Teaching and Learning

The use of LLMs for teaching, of all subjects, and at all levels, falls into two general areas: instructor use of ChatGPT vs student use of ChatGPT. Whereas early discussions centered around “what are our students doing with ChatGPT (and how can we stop them)?” many conversations have moved on to “how can we use these tools to reduce our own workload and improve teaching and learning?”

Another dichotomy to be aware of is using ChatGPT as an instructional tool (for teaching or assessment or both) vs teaching students how to use ChatGPT (including showing them its limitations and how the tool could be used in their industry). It is important to teach students how to use ChatGPT, when not to use it, how to best create prompts, how to critically evaluate its output, how to best leverage it for process goals (thinking, learning) as well as product purposes. However, who should be teaching this? Does every instructor need to become immediately fluent in not only what LLM-based tools can do, but also, what they cannot do, how their use can be detected, and how they are used not only in the academic discipline, but also in industry at large?

Unfortunately, the short answer, at least for now, is probably “Yes.”