This page offers AI-related learning activities for instructors to adapt or reuse in their classes. These activities are designed to foster high-impact educational practices, from critical thinking and synthesis skills to peer learning and metacognition. If the goal is to prepare students to negotiate a world increasingly awash in AI tools and systems, then these activities offer a starting point for instructors here at CUNY.
Before proceeding, instructors should feel comfortable discussing basic AI terms and principles in addition to ethical problems raised by generative AI today. They should also feel confident with walking students through the process of leveraging these tools thoughtfully as part of the activities modeled below.
Instructors should also test these activities before they adapt its design and implement it in their courses. Conducting multiple dry-runs can go a long way toward effective instruction with a technology as dynamic and uncertain as AI. Practicing these activities also carries the added benefit of putting instructors in touch with recent updates to AI tools. Here it is key for instructors to question how these changes impact student learning, especially as AI tools begin to gain traction at CUNY and in education more broadly.
Revision and Peer Review
Student Drafts and AI Peer Review
In Assignment Makeovers in the AI Age, Derek Bruff suggests an exercise in which students use an assignment rubric to evaluate and review the first draft of an assignment by a student volunteer in comparison to a mock draft generated by ChatGPT. Bruff explains further,
I would definitely share ChatGPT’s first draft with my students during class and have the students use the assignment rubric to evaluate it. I’ve often had a student volunteer to share a draft of their essay for peer evaluation using the rubric (and a classroom response system), but ChatGPT’s draft would be a better choice. None of the students would feel the need to hold back on harsh critiques out of fear of hurting the writer’s feelings! And this activity would lead to a very useful discussion about writing and my expectations for the assignment.
This approach frames student participation through an evaluative lens and supports peer-to-peer learning without compromising learning objectives. It also encourages students to distinguish between the nuances of human writing and AI-generated content. Bruff’s activity is especially relevant for those who assign personal or reflective writing projects, such as literacy narratives or autoethnographies. Bonus points for instructors who can draw attention to the wonderful and diverse complexity of CUNY student writers, whose voices are likely to pop off the page compared to the uniform, if not homogenous language of an AI’s output.
Learning Goals
- Comparative Analysis: Develop students’ ability to critically evaluate writing using a rubric, distinguishing between human and AI-generated content.
- Critiquing AI in Writing: Gain insights into the characteristics and limitations of AI-generated writing compared to human expression.
- Reflection on Language Differences: Facilitate engagement with the value of cultural diversity and language differences in CUNY student writing
Suggested tools: ChatGPT, Bing Chat, Claude 2
Brainstorming and Invention
Post-It Note and AI Brainstorming
On an episode of the Intentional Teaching podcast, Garrett Westlake offers an exercise in brainstorming that asks students to take down ideas for a project or assignment on Post-it notes, working within a specific time frame to produce as many brainstormed ideas as they can.
Concurrently, instructors use ChatGPT to generate additional ideas based on the same prompt. This dual approach allows the class to compare and contrast their ideas with those from ChatGPT, enriching the brainstorming process.
By integrating ChatGPT, instructors can extend the scope of brainstorming, uncovering overlapping and unique ideas. This comparison not only serves as a foundation for engaging discussions but also prompts students to reflect on and value the diverse knowledge and perspectives they contribute, highlighting the distinct insights human minds offer compared to artificial intelligence.
Learning Goals
- Brainstorming Practice: Enhance ability to generate a wide range of ideas during brainstorming sessions
- Collaborative Learning: Develop capacity to brainstorm and select ideas in collaboration with peer learners
- AI-Assisted Ideation: Practice using AI text generators for the purpose of invention and brainstorming ideas
Suggested Tools: ChatGPT, Claude 2, Bard, Bing Chat
AI Literacy
Prompt Engineering with ChatGPT
Teaching students how to get the most out of generative AI models can be a daunting prospect for instructors. A starting point can be to review OpenAI’s guide to prompt engineering, which provides strategies and tactics often used to yield better results with ChatGPT. Strategies are broken into lists of tactics that scaffold users through the process of prompting ChatGPT with specific goals and techniques in mind.
OpenAI’s guide can be useful to instructors who feel that prompt engineering can or should be taught as an emergent research practice or compositional method. These strategies and tactics promote students’ functional awareness of ChatGPT and similar LLMs, while also providing clear and well-defined guidelines for how to engage intentionally with AI applications at large.
Learning Goals
- AI Tool Exploration: Develop the skill to explore and experiment with various features and functions of AI tools, understanding how different inputs can affect outputs.
- Effective Prompt Formulation: Learn to craft effective prompts that lead to desired outcomes, understanding the importance of specificity and clarity in communication with AI.
- Adapting to AI Responses: Students will gain the ability to adapt their prompts based on AI responses, refining their approach for improved interaction and results.
Compare and Contrast
When you distribute your syllabus to students, consider also distributing a version of the syllabus crafted by ChatGPT or another generative AI tool. Share the prompts you used to generate the syllabus with your students, and ask them to take five minutes to compare and contrast the two documents, jotting down whatever jumps out at them. Use the following questions to start a discussion:
- How does each document imagine students in the class?
- Where is there overlap in course content, and where is there divergence, and what might these instances suggest about how the class will approach the field of study?
- What can you discern about the professor from each document?
- What can you glean about the institution?
A cheekier approach to this exercise is to distribute the two documents without indicating which was created by generative AI, and asking your students to figure out which one. This approach is adaptable beyond the “syllabus day” moment in a course, and applicable to a variety of documents you might distribute, such as assignment prompts, primary sources, resource guides, and other readings.
Learning Goals
1. Critical Discernment: students will strengthen their capacity to evaluate and assess authorship, perspective, authenticity, and bias in texts.
2. Putting Disciplinary Learning in Context: students will gain experience thinking about how the curriculum for a course within a field takes shape, what choices are made, and why.
