When considering the many new and ever-changing AI tools, their ubiquitous and exponential nature can make it difficult to know where to start and how they might be best used in the classroom. One way to approach Large Language Model (LLM) AIs like Chat GPT and Bing is to think of them as assistants or thought partners. The AIs can assist with instructional development and preparation tasks so that instructors can focus more on the pedagogy and practices that lead to the best outcomes for learners. When the capabilities of AI are paired with what instructors and students envision, the potential for AI to enhance the teaching and learning experience becomes more apparent.
Students need and prefer varied examples, scenarios and case studies when learning new and complex information. Generating these things in order to expand on topics covered in class is one of the best use cases for LLM AIs, particularly because it is possible to seek endless variations to accomplish class objectives.
Example prompt: “Write three detailed scenarios that help demonstrate the accommodating style of conflict management. Each scenario should be 1-2 paragraphs long, written at a university level. Then, write 5 critical thinking questions related to the scenarios that focus on contrasting the accommodating style of conflict management with the other styles.”
Students generally learn best in contexts they already know, which makes providing targeted, varied explanations and comparisons especially effective. Generating focused summaries and explanations to help process long and/or difficult information is something students probably already use AI to do and can be leveraged to include a larger variety of input.
Example prompt: “Explain the kinetic theory of gases in a way that a non-scientist could understand. Frame it as a comparison to something most university students are familiar with in their daily lives. Also add a series of instructions for how to demonstrate the kinetic theory of gases in a simple science experiment.”
Diversity of information and examples is helpful for students, especially in discovering nuance, sparking new ideas and increasing engagement. On the more creative end of the AI spectrum, LLMs can take one set of text and transform it contextually, stylistically or structurally. This might look like explaining something in the voice of a certain person, turning a book into a song or vice versa, or visualizing data sets in multiple ways.
Example prompt: “Write a summary of Plato’s Republic in the form of a Stephen Colbert monologue and also in the style of an Ezra Klein essay. Then write a hip hop song about Plato’s philosopher kings.”
Beyond its ability to support instructors and students in content generation, LLMs can also be used to review content. This application of AI allows for feedback, error identification and assessment. On the more technical end, AIs can help identify errors in code (as well as generate new code) to help build new programs. At the same time, AIs can offer feedback on writing samples, make and grade assessment questions or offer suggestions and improvements on ways to teach and learn specific concepts for specific learner levels. Like any tool, it isn’t perfect, but it can be used to help students and instructors consider potential issues.
Example prompt: “Proofread the following text for spelling and grammatical errors. Provide a corrected version and list the corrections that were made. Offer suggestions to make it more engaging.”
Portions of this toolkit were developed by Western Michigan University's WMUx Teaching and Learning Team and adapted with permission for NIU by the Center for Innovative Teaching and Learning