The Center for Innovative Teaching and Learning (CITL) is rolling out a new ethics module for including AI in your teaching. It will be part of IU's instructor-focused Canvas course, which seeks to provide a pedagogical foundation from which you can plan AI use in the classroom. The following is an excerpt from the upcoming Canvas module on some of the ethical considerations surrounding AI use. (If you're looking for in-depth training on a range of GenAI skills, be sure to register for IU's self-paced GenAI 101 course.)
Along with other ethical concerns regarding generative AI use, you and your students may have legitimate concerns about how much water and energy AI uses. Research by Maximilian Dauner and Gudrun Socher indicates that LLMs can use substantially more energy than previous methods. Studies have demonstrated that AI produces modest productivity gains (2.8% in this study), but uses exponentially more amounts of water and energy to produce those gains.
In order to better quantify these concerns, Professor of New Media Jon Ippolito has created a "calculator" that generates numbers based on the averages of the amount of energy consumed by performing various digital tasks, such as interacting with generative AI (text- or image-based), doing a Google search, storing files online, or charging your cellphone. You can play with it to see the energy costs associated with these common digital tasks for yourself.
After your own explorations you might consider designing a classroom activity. By letting students explore and educate themselves on these costs, you can help them make informed choices about when and how they use generative AI. Here are the basic steps to the activity:
- Students compare an AI task with something they routinely do (like charging their cellphone).
- You ask them to find equivalences to evaluate AI uses such as, "How many AI text searches are equal to charging a cellphone?"
- You conclude by facilitating a class discussion about these comparisons.
Building on that discussion, Jon Ippolito and Tai Munro offer a structured activity that guides students through creating personal guidelines for responsible AI use based on using what they learned.
IU instructor Miranda Rodak engages her business writing students in conversations about the energy consumption of generative AI through her "Hire AI" assignment, a scaffolded, active-learning framework that spans the semester. In this assignment, students "hire" an AI assistant to support their learning, but they are responsible for making informed decisions about when and how to use it.
Early in the course, students examine ethical and practical implications of AI—including environmental impact—so they can define thoughtful parameters for responsible use. Rodak asks: when does it make sense to involve AI, and when does it not? She uses an analogy: if students have a car, they could drive anywhere, but they likely wouldn't drive next door because that would waste gas and energy. Similarly, is the energy required for AI to write an email worth the benefit?
One final consideration: Keeping in mind the recommendations from the American Association of University Professors that students will likely need AI literacy skills to be competitive in the job market when they graduate, how much should you, as their instructor, require or encourage them to (not) use AI?
The ethics module will also cover accessibility, biases, and the potential homogenizing effects of AI on writing style—as well as generative AI's potential to offer personalized learning. Look for the rollout of this final course module in the coming days.
By exploring these ethical and pedagogical dimensions, you will be better prepared to help students make thoughtful, informed choices about AI use. We at the teaching centers are also always happy to meet with instructors and have a conversation about whether and where you can best use AI in your courses. Contact us to schedule a consultation today!
About the author: Eric Brinkman, PhD
As an Instructional Technology Consultant, Eric is dedicated to supporting and collaborating with instructors to implement tools that enhance equitable, student-focused teaching. Currently, they focus on developing materials to help instructors integrate generative AI into their teaching practices. Eric offers workshops, facilitates a Faculty Learning Community, and has co-designed a set of Canvas modules to assist instructors in getting started using AI in the classroom. Learn more about Eric here.