If we put a quarter in a jar every time we heard the term ‘ChatGPT’ used in the last year, we’d be pretty rich. We suspect you would be, too. But as Dr. Derek Bruff—author and host of the Intentional Tech podcast—reminds us, the rise of AI may not be such a detrimental thing in the classroom. In fact, it may offer the nudge we need to rethink assignments for the better. 

At our recent webinar, Bruff offered suggestions on how to reframe learning objectives in light of AI. He also left instructors with simple adjustments to make in their own courses that maximize (or minimize, depending on level of trepidation) the use of AI.

→Recording: Dr. Derek Bruff on how to incorporate AI into assignments

Why might we need to change our learning objectives in light of generative AI?

It may seem like a massive undertaking at first. But AI can help us shift the focus from rote memorization to application. As John Warner, former college instructor and author of Why They Can’t Write, argues, we may need to evolve away from rewarding “surface-level competence” when grading papers to activities that encourage reflection, critical thinking, and synthesizing information. 

Imagine you teach a computer science course. Perhaps you move from asking students to generate code—something that ChatGPT can easily do—and instead ask students to evaluate code written by their peers. One of the major concerns with AI is that it can easily shortcut student learning. If you’re teaching an upper-year biology course where students are required to identify different species of birds, you might prevent the use of AI in order to help learners develop the foundational skills they will need as future biologists. Because AI tools are often inaccurate, you might explain that you’re choosing to minimize their use until students have a good command of the fundamentals in order to make them more discerning users of these tools in the future.

What are some ethical ways for students to use AI during the assignment process?

  1. Put words on a page: We know filling a blank screen with prose can be the hardest part. My colleague at the University of Mississippi allows students to generate an introduction for their essays using AI. Some choose to keep their generated response, while others re-work what the computer just gave them.
  2. Make sense of complex readings: Let’s be honest, not all of us read an entire journal article from start to finish. So there’s a good chance students don’t do this either. You might let students upload a research article into an AI platform and request a 500-word summary to demystify complex jargon or technical language.
  3. Seek targeted feedback: AI can guide students when forming first drafts of essays or business cases. Specificity is the key here. Ask students to feed their document to AI and use a prompt like, ‘evaluate the evidence used to support the main argument: is the evidence relevant, sufficient and effectively integrated into the argument?’ You might go a step further by getting students to write a one paragraph reflection at the end of the assignment asking them what they’ve learned from interacting with AI, what they’re choosing to change about their paper and what elements they’re keeping in their final draft.

How might educators benefit from using AI when forming assignments?

AI is great for reading the room. I’m a big fan of classroom response systems. But the challenge is, if you pose an open response question, it’s hard to get a sense of hundreds of student comments at a glance. That’s where AI comes in. After facilitating an online discussion, you might feed these responses into ChatGPT and request a summary of the main trends. You’ll receive an instant look at what the key patterns are in students’ responses and can use this information to tailor your lectures or future quizzes accordingly. It can also be a compass of sorts that guides you in making strategic assignment choices based on potential learning gaps in your course.

What are some effective assignment ideas that foster critical thinking and evaluation in our AI world?

The permanence of Generative AI doesn’t mean you need to overhaul your entire course structure. It’s about rethinking assignments and the role AI might play in them. I’ll borrow an example from a statistics course I previously taught. Students were asked to find a data set and complete some basic statistical analysis, then present their results in an engaging infographic. There’s both analytical and creative components at play—which typically appear in almost every course no matter the discipline. Here’s how AI might guide struggling students.

Tools like ChatGPT can create visual scatter plots for large data sets. Google Gemini is good at the reverse. You might ask students to input an image of data visualization and ask the tool to explain the trends in the graph. I’ve also found success using Claude to identify the most important information to showcase in infographics—like breaking down pizza prices in your local county relative to income level. If a student is struggling to get started on their project, AI can provide a very practical roadmap of how to begin. Again, as John Warner advocates, be sure to review your rubric to avoid allocating points to work that’s now too easy given widespread access to Generative AI tools like the ones shared here.

How do you suggest giving assignments a ‘makeover’ to let students learn with or without AI?

There’s a five-step process we can all follow, no matter the discipline.

  1. Why does this assignment make sense for this course? Evaluate how the assignment is critical in helping students prepare for their upper-year classes or careers.
  2. What are the specific learning objectives? Start by plotting out your objectives before thinking about your specific assignments. Then, consider how AI tools might intersect, detract or enhance the learning experience.
  3. How might students use AI when working on an assignment? Spend time experimenting with different AI tools and see what the possibilities are. Once you’ve familiarized yourself with the affordance of AI, you’ll better understand how to anticipate student usage.
  4. How might AI undercut the goals of your assignment? Determine how students might use AI to jump from point A to point B. Again, you might consider revising your rubrics to ensure you’re allocating points towards application and not just completion or rote memorization.
  5. How might AI enhance assignments and where would students need help figuring that out? This final recommendation comes from Anna Mills, English Instructor at the College of Marin. The more authentic your assignment is, the more supported students will feel. Here’s the most rewarding part: students are less likely to find shortcuts in your assignment if they feel they’re fully invested and know they’re getting something out of it.

→ Recording: How to improve assignments with (or without) AI

Tagged as:

, ,