University courses are notoriously reading- and writing-heavy, so it’s no surprise that EAP courses have a similar setup. And when there’s a text to be read or written, large language models can confidently barge in, as these are the tasks they are the most apt at solving. I asked a very good colleague and friend of mine, Anna Csíky, who’s currently teaching such a course, to tell me how she had approached the question of AI in EAP at the beginning of the term.
I also teach a couple of courses at university level, so something that’s been on my mind since the moment artificial intelligence appeared on the scene was how this is going to affect my teaching, especially the assignments. This is the first academic year when AI is not a curious novelty anymore but an additional player to be dealt with. So I asked myself the question: If most of an EAP course involves reading and writing texts and if an AI tool can swiftly complete these, then what’s the whole point of taking such classes, and how will essays in EAP courses change? I interviewed Anna Csíky to see what she thinks about the question and how she sees the future of EAP.
Jo Szoke: What was your first thought when you learned about the capabilities of AI and you knew you were going to teach an EAP class?
Anna Csíky: Actually, it was the other way around! I already knew about AI tools, like ChatGPT and Google Bard, and I was intrigued by the implications of students having access to such powerful aids. Why would anyone want to write an essay themselves when a machine can do it for them in a minute? So, I thought to myself that the best way to get to the core of this issue is to test it in the most authentic environment, so I decided it was time for me to return to the EAP classroom.
Jo Szoke: So, what kind of a role did you give to AI on your course?
Anna Csíky: To me, an EAP course is primarily about helping students communicate their ideas with more precision in an academic context. AI does not change that, but makes it more tempting to focus on giving students a fish rather than teaching them to fish. In an EAP class, fishing is the subskills we are developing, such as expressing your point, formulating a counter argument, or deciphering the key message when faced with a complex text.
My aim is to create a learning environment where students understand what subskill we are working on, and for them to see that they get more out of engaging with the task than simply getting a robot to do it for them. Practical ideas that have helped with this are explicitly focusing on the process rather than the product by, for example, having multiple drafts of written texts or setting scaffolded and collaborative speaking tasks, so students can learn between different stages and from each other, improving their skills organically. AI can be used throughout the process but transparently, sharing and reflecting on what students learnt from each LLM conversation.
Jo Szoke: Did any of the assignments have to be changed?
Anna Csíky: Yes! For example, this course used to have a summary task, which asked students to create a written summary of a longer text. Now this is a great example of a task that can so easily be done with or even by AI, so one idea was to change the modality, with students creating video summaries instead. But then me and the other teachers teaching this subject realised that the script for the video can also be generated and then the student just reads it out with the help of a teleprompter!
So, again we dug deeper to find a human skill worth developing and decided that the unique thing students can add to this task is their own voice and opinion. So, now they record a video summary but make it personal by adding lots of commentary and reflection. AI can do parts of this, of course, in which case we ask them to attach screenshots / provide links to chats, and students are prompted (pun intended 🙂) to be actively involved in the process.
Jo Szoke: What do you think your students’ relationship is with AI? Do you think they want to use it to cheat or to make their lives easier?
Anna Csíky: That’s an interesting question. I know I use it a lot to make my life easier – but I also like to think that I learn from every exchange I have with it. So I’m trying to model and encourage that approach. A student came up to me the other day to ask if they could use a paraphrasing site for their essay. I suppose they wanted to make sure their English was perfect – which is great, and I appreciate that they asked me, but as I said, I want AI use to be transparent and to focus on the learning process rather than the product. So they can also submit a record of what they’ve learned from comparing their version with the paraphrased one.
I also believe it is better to ask for specific sections or sentences to be paraphrased, and not the entire essay. What’s important here is having the right mindset: do I want to tick a box, or do I want to become a better writer? How can I make sure that the input I get from AI becomes intake and next time I won’t make the same type of mistakes?
I wholeheartedly agree with Anna, especially with regard to her final point. As EAP teachers, we can help our students find the sub-tasks that AI can do for them during the learning or research process, but at the same time, we need to emphasise the human elements that increase the value of the task and ensure its originality. This can include commentaries, reflections, or corrections to the generated material.
Discover a world of resources relating to EAP, including blogs such as “Is That Your Final Draft?! 3 Tips for Teaching Editing” and “Planning an EAP course for academics“.