Generative AI
for Teaching and Learning

Rapidly evolving generative AI tools include text-generating AI chat services (e.g., ChatGPT, Bard, Claude), as well as image-generating, audio, and video AI tools. These machine learning-powered tools are increasingly integrated into our daily communication apps, word-processing software, and technology broadly. In order to best serve our students, we need to engage with generative AI in meaningful ways across the curriculum.

At one level, generative AI tools are no different than any other tools with regard to their impact on teaching decisions. Faculty will need to consider how to leverage tools to benefit student learning, asking question such as:

  1. How can this help me prepare course materials or handle administrative aspects of instruction?
  2. How, if at all, can they help students master my course outcomes?
  3. Where, if at all, in my course should I be explicitly teaching how to use the tool?
  4. Where, if at all, is it appropriate in my course for students to use the tool independent of goals one and two?

Faculty also need to be aware of pitfalls and dangers, such as:

  1. How might use of the tool negatively impact students’ learning in the course?
  2. What equity and access issues does the existence of the tool raised for my course?
  3. How will I address concerns with data privacy breaches, intellectual property protection, algorithmic biases, and ”hallucinations”, situations where generative AI provides false information?

Ultimately, individual faculty will need to make decisions based on the context of their course, course objectives, students’ academic progression, and disciplinary-specific goals of their students’ learning experiences.

One of the challenges faculty have in making these decisions is that the tools are potentially moving faster than research into the use and impact of the tools. However, UCI has significant expertise in this space, and we will continue to update the information as we learn more from research. The following links provide important resources in this space.


Links to UCI statements and materials

A Statement from UCI Writing Pedagogy Experts

Teaching writing and communication, including digital tools, depends on context. Instructors and curriculum designers must consider students’ needs, disciplines, and academic progression. We assert the following:

  1. Students deserve quality writing and communication instruction, which promotes critical and creative thought, as well as good writing and communication habits.
  2. Students should learn about various tools – digital and analog – so that they better understand the capabilities, limitations, creators, purposes, and societal + environmental impact of these tools.

The rest of this document provides additional information for getting started and a range of resources.

Where to Start?

Resources

Advisory Group

Where to Start?

Experiment with at least one of the generative AI tools 

  • Try something personal: “Suggest 5 recipes for dinner tonight using chicken and broccoli” Stir fry it is!
  • Try something for your class: “What are 5 different metaphors for explaining logarithms to undergraduate students in math class?” Hmmmm, maybe the password decoder?
  • Try out one of your assignments to see what generative AI would produce: “Write a 5 word poem on what you did this summer.” I’m a machine; I pondered.

Need help? See “Setting up ChatGPT


Explicitly and transparently address the use of AI in your class in your syllabus and in your initial class meeting.

  • Need some ideas? This blog has links to examples.
  • For information about disabilities, see the UCI DSC.

Ideas to get you started:

As with any set of recommendations and resources, these are not meant to be blanket recommendations that apply everywhere (even if some external resources may appear to be phrased that way). Faculty should use their own judgement as appropriate.

 

Prompting Generative AI: Tips


  • This article and video provide some useful background information on how to instruct ChatGPT in order to get useful, relevant content.

One thing to consider before you use generative AI… verifying accuracy is often a first step before using AI in any application …

Note: Keep in mind that safe with respect to accuracy does not necessarily mean you should use it!
Adapted from ChatGPT and Artificial Intelligence in higher education.

An example from the Vice Provost


There is a difference between “what ChatGPT can do and its advantage as a tool” and “how ChatGPT can help advance the learning experience of a particular class”. I think this nuance is important because at least in the short term (if not long term), faculty will continue to use the tool differently in their courses, so we need students to be alert and understand this. Also, our goal is to move faculty to maximize the use of ChatGPT to help learning AND to help students learn how best to use ChatGPT – which are two different things.

Concrete example from Mathematica and a math methods course I taught (I like this example because we did both in the same course):

In the first part of the course, when we reviewed a particular math concept (such as eigenvectors) the students could use Mathematica to do the algebra but could not use the “eigenvector commands” because they were learning the algorithm to do eigenvectors and using mathematic to do algebra helped them learn the concepts by focusing on the algorithm and not the algebra, but by the end of the course, for most eigenvector problems they were using the command so they knew how to take advantage of the tool to do what it was designed to do.

Resources

Do you need materials to support students’ understanding of generative AI and how it works?


  • This instructor resource has suggested reading and a video you could assign to students or you can use this student handout.
  • This detailed explanation by Stephen Wolfram is a great technical resource.
  • This statement by the MLA-CCCC Joint Task Force on Writing and AI provides additional context for writing instructors. This UNESCO guide is a useful general reference.

Would you like to provide students with materials on generative AI’s inherent limitations and biases to make them better users of AI in your class and in the world?


  • Here is an instructor guide with suggested texts, videos, and deeper dives, as well as a student-facing resource.
  • For more information on hallucinations, you can reference this article and video.

How to Cite Generative AI for Academic Work


OVPTL Generative AI Advisory Group

The OVPTL Generative AI Advisory Group was established in Spring 2023 with the mission of providing guidance to the UCI faculty and campus about approaches toward ethical and effective uses of generative AI in higher education.

Mark Warschauer

Professor of Education [chair]

Tom Andriola

Vice Chancellor for Information Technology

Qian Du

Associate Director of Global Languages and Communication

Daniel M. Gross

Campus Writing and Communication Coordinator

Amalia Herrmann

Lecturer and Digital Pedagogy Coordinator, Humanities Core Program

Rikke Ogawa

Assistant Librarian for Public Services

Megan Peters

Associate Professor, Cognitive Sciences

Brad Queen

Director of Composition Program

Daniel Ritchie

AI in Education NSF Graduate Research Fellow & Chair of AIforCA

Brian Sato

Director of Teaching Excellence and Innovation

Tamara Tate

Associate Director of the Digital Learning Lab

Waverly Tseng

Graduate Student Researcher on Generative AI in STEM Writing

Michael Dennin

Vice Chancellor for Teaching and Learning [ex officio]

Gillian Hayes

Vice Provost for Graduate Education, Dean of Graduate Division

 This material is based upon work supported by the National Science Foundation under Grant No. 23152984.  

© 2023 The Regents of the University of California