At McMaster, the Provisional Guidelines on Generative AI in Teaching and Learning ask that if you do use generative AI in your teaching materials or assessment practices that you share this with your students both in the course outline and in class. Sharing your use of generative AI with your students is intended to build trust and transparency, and to acknowledge that you are also using – and learning about – generative AI.
Those same Guidelines suggest that you can use generative AI with data collection turned off to provide students with formative feedback on assessments. Formative feedback is feedback that is not for grades, but rather gives students fast and specific advice on how to improve. Formative feedback from a generative AI tool might be given on an essay outline or draft, for instance, while you or the teaching assistant would be responsible for assessing and grading the final essay submission.
Finally, the Guidelines also ask that you check the accuracy of any AI-created content. Recognizing that these tools “hallucinate” – or come up with factually incorrect responses – it is important that you check the accuracy of any content you might use in class, or any feedback offered to a student.
With that said, here are some broad categories where generative AI may be useful to you as an instructor:
Generating Test Questions and Assignments
By prompting a generative AI tool with the specific context of your course, as well as the subject you are aiming to assess and the kind of question or assignment you are interested in, the generative AI tool can offer many – many – examples of test questions at different levels of complexity, or different types of assignments. You can even ask for assignment ideas that meet the criteria of authentic assessment discussed in the chapter on assessment, or for assignment ideas that incorporate pedagogical approaches you value (e.g. problem based learning, community engaged learning or case based learning).
Generating Examples, Explanations and Counter Positions
Students benefit from practicing what they are learning with examples. Many, many examples. Generative AI can be powerful in producing lots of examples for students to practice with, while also providing students with feedback on whether their submitted responses are correct, or how they might improve on a response. This personalized, immediate feedback is incredibly powerful for learning.
It can be challenging sometimes to describe a concept at many different levels of complexity. Some courses – especially those with no prerequisites – may have a range of experience and abilities in the class. Using generative AI tools you can quickly develop (and then check for accuracy) multiple explanations for a course concept. You could even have these explanations be written in unique and memorable ways – like, explain the carbon cycle in a limerick or describe the Canadian political parties as characters on the Simpsons.
Generative AI tools like ChatGPT can take on different personas by prompting – for instance, you could ask the tool to “pretend you are a heart surgeon” or “act like you are the Prime Minister”. In assigning this persona, the generative AI tool will produce text written as if from that position. This kind of role can be useful in inviting unique perspectives into a class discussion, or providing a provocative counter point.
Gathering Ideas for Class Activities and Assessments
Confronted with the challenge of generative AI you may be looking for new ways to teach a concept or skill, or new ways to assess a learning outcome. Generative AI can provide customized suggestions for interactive and engaging classroom activities (e.g. suggest six different interactive ways I could teach an auto-ethnographic research method to a third year, online class of 60 students in Sociology), as well as assessments that either incorporate generative AI or make generative AI less likely to be used.