Gen.AI can be of enormous help in explaining vocabulary. There are various things we can do as learners to explore, understand and get exposure to lexical items, such as asking for definitions, requesting example sentences with the target words, or asking Gen.AI to write stories incorporating that vocabulary. In addition to these methods, teachers can also ask Gen.AI to generate conversation questions incorporating the target words for further exposure and practice in the language classroom. In this blog post I am going to give this last method a twist.
How about having the learners think of questions for Gen.AI to provide answers to? By doing so, we are giving learners opportunities to engage with the target vocabulary in different ways, which increases the chances that the language will stick in their minds. There is also an added element of interactivity and modality by having students listen to the answers provided.
Take these tasks from Open World C1, Unit 2, page 29 (Cambridge University Press). You can download a free sample from the publisher HERE.
Task 2. Put the following words and phrases into the most suitable transport categories in the Venn diagram. Use a dictionary if you have one. track (n) reverse (v) boot (n) give way (v) carriage (n) board (v) cabin (n) shuttle (n) congestion (n) wreck (n) overtake (v) commute (v) divert (v)
In task 4 students have to “decide which of the words in Exercise 2 have other uses/meanings in different contexts and as different parts of speech” (as in, for instance, keep track of things or to turn into a nervous wreck or to be fully on board with an idea).
Let’s focus our attention on these three lexical items. In addition to having students do some or all of the tasks from the coursebook, we could also ask them to pick a lexical item and come up with an interesting conversation question for ChatGPT. We will use the ChatGPT voice extension or the app for Android or iOS devices to read out some of the questions written by the students. It’s important to think of a role for ChatGPT, asking it to act as a certain type of person and keep the answers short. Here are three different prompts based on imaginary questions written by the students, using some of the target vocabulary.
Prompt: Imagine you are an 80 year-old man. What’s one thing you struggle to keep track of in your daily life? Your answer should take about 15 seconds to read.
One thing I struggle to keep track of these days is where I put my glasses. It feels like I’m constantly setting them down somewhere and then wandering around the house trying to remember where. It’s funny, really—half the time, they’re right on top of my head!
Prompt: Imagine you are a parent. How do you handle your kids when they turn into a nervous wreck over small things? Your answer should take about 15 seconds to read.
When my kids turn into a nervous wreck over small things, I try to stay calm and reassure them. I listen to what’s bothering them, acknowledge their feelings, and help them see things from a different perspective. Sometimes, a little humor helps lighten the mood, and they realize it’s not as big as they first thought.
Prompt: Imagine you are a teacher. Have you ever been in a situation where you weren’t fully on board with an idea but had to go along with it? Your answer should take about 15 seconds to read.
Yes, I once had to implement a new grading policy that I felt wasn’t entirely fair to students. I didn’t agree with it, but I followed it while giving extra support to those who might struggle. I also shared my concerns with the administration, hoping for future improvements.
So we would hide these texts from view and have our students listen to the voice version. This listening task doesn’t eliminate the option for students to discuss those questions in small groups or as a whole class. While some of ChatGPT’s responses may come across as somewhat stilted, the listening component adds extra value, making this practice worth exploring.
AI in ELT 