How ChatGPT Will Change Language Teaching

March 2023

This month’s issue has been co-published with the JALT CALL SIG. By pairing their tech-savviness with our brain-nerdiness–in addition to the input from our readers–we’ve put together a deep dive into Generative AI technology. While we focus in particular on the ChatGPT application, we do take a broader look at other forms of AI and what this improving technology means for teachers, students, and language learning. 

Our cover: “ChatGPT is scary good. We are not far from dangerously strong AI.” – Elon Musk

Cover photo by Pavel Danilyuk on Pexels; others from Pixabay, Openverse, and Unsplash

Watch before you read...

Since its release late last year, the large language model (LLM) ChatGPT has garnered considerable attention from educators. In this issue on the intersection of ChatGPT and foreign language learning, we teamed up with JALTCALL, the JALT special interest group for Computer-Assisted Language Learning. Looking at the big picture, the Lite video shines a spotlight on fundamental questions concerning human-AI interaction. In the Main video, computational linguist Emily M. Bender and cognitive scientist Evelina Fedorenko discuss human communication and LLMs. Setting the stage for our Think Tank, Mohammad Khari gives an explanation of ChatGPT and summarizes this month’s videos.

In the Think Tank, Louise Ohashi examines how the judicious use of AI can support foreign language learners. Just after that, Glenn Magee explores predictive language processing in humans and AI. Bob Cvitkovic emphasizes how our choices are essential in the face of rapid tech development and an uncertain future. Next, our readers share a wide range of thought-provoking ideas, followed by an invitation to five AI sites. Then, Phil Norton comments on the output ChatGPT generated when prompted to advise foreign language learners and teachers. Shifting our focus to neuroscience, Ken Purnell, Skye Playsted, and Justin Kennedy delve into how AI can impact neuroplasticity, lesson preparation, and assessment practices, while in our PLUS, Laura Gibbs and Heather Kretschmer return to share more creative ideas for short writing tasks.

Our Thoughts on ChatGPT

ChatGPT: The Double-Edged Sword of AI Language Models Mohammad Khari

Do you prefer having all your emails in one place or do you think it is better they are sorted by someone/something else and some sent to your spam folder? Do you enjoy the music or movie recommendations on streaming platforms like Spotify or Netflix? Have you ever expanded your network on LinkedIn following the accounts and people suggested to you because they were similar to your interests? The list of daily applications of AI is endless and we seem not to be bothered by most of it. We find it useful and some say it is a necessary evil. The same is true about the more specific use of AI in relation to language. Most people are fine with their text being checked by apps for spelling and grammar. So, what is it that fundamentally differs when it comes to Large Language Models (LLMs) based on transformer architecture? Aren’t they just another form of AI?

Think Tank Articles

Rethinking Education in the Age of AI Louise Ohashi

Startling headlines like this from the New York Times (Huang, 2023, January 16)[1] have been widespread since ChatGPT’s release, but how is the average teacher responding? I’m currently analysing data from a worldwide study with 368 language teachers to find out (check Twitter for updates), but in the meantime a quick search online shows reactions at both ends of the spectrum. On one side, innovators have already created learning materials and set AI-integrated assignments. On the other, ChatGPT has been banned and all written work moved to class time to prevent cheating. Love it or hate it, ChatGPT cannot be ignored because we are in a new Age of AI. The dial is about to be turned up.

Similarities and Differences between Predictive Language Processing in Human Brains and Machine Learning Systems such as GPT-3 Glenn Magee

Predictive language processing is how our brains predict the meaning of language we hear or read. The theory is that our brains constantly create mental models to help us predict things, such as the next word in a sentence, based on our past experiences and knowledge. That is important because we do these things in real-time without having to stop and consciously consider the meaning of every word or sentence we hear. Our brain processes are automatic and interact with other brain functions to quickly and efficiently grasp the underlying purposes of words, sentences, and texts. Many Artificial Intelligence (AI) systems also use predictive language processing to predict what word or words will come next in a sentence or text. Artificial intelligence aims to provide software that can explain output based on input. This is important because any AI system’s crux is the input we are giving it. In other words, the quality of the data or information fed into artificial intelligence systems determines the outcome.

AI is on Track to Take us One Billion Meters from Today Bob Cvitkovic

I fear that the genie is out of the bottle, the cat is out of the bag, and the toothpaste cannot be put back in the tube. Pick your favourite idiom but, as of November 30, 2022, the paradigm has shifted, and we are in the middle of it. As I write these words in the cloud, Google Docs instantly saves all my changes, the spell-checker auto-corrects every word, Grammarly makes sure I am following all the proper rules of language use, and QuillBot suggests synonyms and rewrites, instantaneously and seamlessly. It is getting difficult to imagine what it was like not to use these tools—and that is what they are: tools, tools for writing. The ideas are coming from my head, and these writing tools notify me when I am breaking the rules of composition and mechanics. Well, that’s not entirely true. QuillBot’s thesaurus and its “rewrite” function tread a gray zone, but at least it is rewriting an idea that I have already typed on the screen.

A Tsunami of Ideas from our Readers Think Tank Readers

Editors: At our request, a number of language teachers sent us their comments on ChatGPT. Most are positive. Most include good classroom applications. All are insightful. But what else could we have expected? Anyone who reads a magazine on connecting brain studies to their profession is bound to be ahead of the herd!

Playing in the Digital Tool Sandbox Think Tank Team

If you’ve been delving deeply into the world of ChatGPT & Co., we suggest taking a brain break. Perhaps go outside and enjoy the last vestiges of summer (or winter, depending on where you live). Maybe have a conversation with another human being. In any case, we recommend disconnecting from devices for a while. When you return, we invite you to play around with these five digital tools. At the time of writing, they are easy to experiment with, free on the web, and don’t require registration.

ChatGPT’s Approach to Teaching: An Interview with the Chatbot Phil Norton

Despite reading about ChatGPT on sites such as Wired and seeing it plastered throughout my newsfeed, I hadn’t really felt the potential impact of the chatbot until I listened to an episode of a Greymatter podcast. Greylock general partner Reid Hoffman interviewed ChatGPT about ChatGPT (Hoffman, 2023a). The interview was part of a series called “Fireside Chatbots,” which all follow the same format: a conversation with the chatbot about its impact and potential. The natural tone and deep insights ChatGPT displayed left me gobsmacked. I thought I would take a similar approach and interview the ChatGPT bot about how it thought it could be used in EFL settings.

Using Chatbots Wisely in Education Ken Purnell, Skye Playsted, and Justin Kennedy

ChatGPT and Microsoft Bing headlines abound on mainstream and social media. Such AI chatbots will change the way we teach, assess, and work. Indeed, the “genie is out of the bottle” and so we must learn how to use it effectively and wisely. Some recent examples from mainstream media include:

“The teacher is dead. Resistance is futile. What sets ChatGPT apart is its exceptional performance and potential to disrupt education. Since its launch, there has been a plethora of online discussion from education professionals sharing examples and opinions about how students will use ChatGPT to cheat on assessments, such as writing essays or coding programs, and how AI could soon replace the teacher.” (McMinn, 2023)

Think Tank Plus

Reusing Limited Resources: Creative Constraints in Writing, Part 2 Laura Gibbs & Heather Kretschmer

This lovely (non-ELT) challenge involves reusing something familiar and quite ordinary: paper clips. Ever straighten out and reshape a paper clip yourself, maybe out of boredom during class? Well, this time it’s encouraged by the teacher. Retired middle school teacher Sheri Edwards writes that she gave her students breaks to take “Apollo Challenges,” where she furnished students with a limited resource and a small task, like: what animal, besides a snake, can you make out of five paper clips? 

Call for Contributions: Ideas and Articles Think Tank Staff

Become a Think Tank star! Here are some of the future issue topics we are thinking about. Would you, or anyone you know, like to write about any of these? Or is there another topic you’d like to recommend? Do you have any suggestions for lead-in, or just plain interesting, videos? How about writing a book review? Or sending us a story about your experiences? Contact us.

Last Week Tonight 

Witty observations on AI from you-know-who,
and he hasn’t even read this issue yet.

We all need to develop the skill of identifying bot-made content. You can practice here!

You against the machine: Can you spot which image was created by A.I.?

Let ChatGPT into classrooms to prepare teachers and students for the future

While there are valid concerns about the use of ChatGPT, it offers positive teaching and learning opportunities for both educators and students. An AI-assisted writing classroom could have many different entry points to broaden student engagement and prepare them for a future that is already here.

http://www.scmp.com

MBE Logo

The MindBrained Think Tanks+

is produced by the Japan Association for Language Teaching (JALT) Mind, Brain, and Education Special Interest Group (BRAIN SIG). Kyoto, Japan. (ISSN 2434-1002)

Editorial Staff

Stephen M. Ryan                Julia Daley                   Marc Helgesen

             Heather Kretschmer          Curtis H. Kelly            Skye Playsted               

    Jason Walters                               Mohammad Khari

 

 

Subscribe

Get issues delivered to your inbox at the start of each month!

Unsubscribe

Stop receiving issues in your inbox.

Reader's Survey

Take a short survey to help us learn more about our readers!