The University of Chicago recently announced efforts to roll out a new generative AI tool that will give all members of the university community access to the latest AI models. Students (and instructors) may see this as the university issuing a seal of approval on the use of generative AI in teaching and learning, potentially leading to more use on class assignments.
In response to this, it is crucial that instructors have at least some basic AI literacy: a working understanding of these tools, their strengths and weaknesses, and the benefits and challenges they can bring to language classrooms. This is the first of a series of blog posts, videos, and workshops to that end.
This post helps instructors understand the basic terminology of generative AI, along with basic guardrails on AI use. Future posts will focus on language teaching more specifically.
Generative AI Basics
First, watch this mini-lecture from Nick Swinehart, Managing Director of Instructional Technology for the Chicago Language Center:
- Generative AI in a Nutshell – Henrik Kniberg (software developer, former Chief Technology Officer for various companies, writer) gives an illustrated overview of generative AI. (Youtube, 17 mins)
- Practical AI for Instructors and Students Part 1: Introduction to AI for Teachers and Students – Part 1 of a series from faculty at Wharton (Youtube, 10 mins)
- Why AI is Incredibly Smart and Shockingly Stupid – TED Talk from Yejin Choi, professor of computer science at the University of Washington. Keep in mind that although the video is just a year old, LLMs are improving very rapidly and some of her examples are outdated in that the latest models can perform them accurately. But her description of why these tools are “shockingly stupid” is still relevant. (Youtube, 16 mins)
Basic Guardrails on AI Use
The next video discusses some basic guardrails regarding privacy, quality control, and academic integrity when using generative AI:
Additional resources from UChicago:
- Generative AI Guidance – IT Services
- How to Think About Generative AI in Your Teaching – Academic Tech Solutions
How to Use Generative AI
How (or whether) to use generative AI in your teaching is a complex question. As a general purpose technology there are infinite ways it could be used, and each person is going to have different views on what constitutes appropriate use for teachers and learners. While I don’t believe all teachers “should” use generative AI, I do think they should understand what it is capable of in order to make informed decisions, and the best way to do that is through hands-on experimentation.
UChicago instructors: I’ll be hosting a workshop titled ‘Getting Acquainted with Generative AI’ on Wednesday, September 11, 1-2pm; you should have received information via email. As the title indicates, this intended for people with no/minimal experience using generative AI, but experienced users are welcome to attend. If you are unable to attend at that time or need the Zoom information, please send me (Nick Swinehart, nswinehart@uchicago.edu) an email.
Below is a brief demonstration of an 8th-grade English teacher using ChatGPT for lesson planning. Again, this is a complex topic on which instructors are likely to have differing views; I only include this video here for a technical demo of using ChatGPT and how some teachers find it useful. This particular example shows how ChatGPT can be used to create materials for very specific contexts, and how the teacher keeps refining his prompts until he gets output he’s satisfied with.
Where to go from here?
Again, future posts and videos will talk more language teaching specifically, and how to approach generative AI in your courses. In the meantime, I recommend experimenting with generative AI either in my workshop, a one-on-one session, or using the countless resources available on the web (e.g., 50 ChatGPT Prompts for Teachers).