Simone Hackett is a Senior Lecturer and Researcher at The Hague University of Applied Sciences (THUAS)
She’s been a keen COIL/VE practitioner over many years, delivering, designing, and coordinating COIL.
Furthermore, she has a lot of experience in conducting research on the topic. Especially testing the effectiveness of virtual exchange in helping students develop or further develop intercultural competence.
Simone takes an interest in looking at the design of coil/VE courses and assignments to see which ones are more effective.
Her colleague is Mark Dawson, currently based in Berlin and working at CODE University of Applied Sciences as a lecturer.
He focuses on student experiences of virtual exchange and how they use generative AI to support collaborative learning.
Recently, he completed his PhD looking at COIL/VE, particularly the student and staff experience during virtual exchanges.
“I’m really interested in this idea of collaboration and how AI might be impacting this collaboration. I think it’s a really fascinating topic and a topical one for our times.”
The Positive impacts of AI on COIL/VE for students
Mark explains how he was part of a research project at CODE that was funded by the German government. It looked at student and staff experiences of facilitating programming learning using online tools.
What surprised them most was the speed of how AI tools were being integrated into students’ learning experiences and processes during virtual exchange projects.
“In a good way, students who had previously struggled with imposter syndrome and were scared to ask for help in class or to their tutor for fear of being judged by their peers and the staff, were suddenly given access to this tool that could explain things in a really accessible and user-friendly way in real time. They could type out something they didn’t understand and quickly get to grips with it without anyone else in the class realising.”
Mark considers this really powerful for virtual exchanges as it enables students to feel more able to engage with the material that was being discussed.
“This is suddenly leveling the playing field for lots of people who previously might have felt unable to engage with normal classroom activities. Some might be very good at speaking English. Others less so. And of course, having a chat GPT available very quickly to translate words or explain something for the ones who are not so good in English is such a nice scaffolding making it a less scary experience to enter these COIL discussions.”
Simone and Mark agree that using AI tools to help with translation helps redress perceived power imbalances in VE, giving all students wider access to LLM chatbots and therefore a broader range of knowledge than traditional institutional databases.
The Positive impacts of AI on COIL/Virtual Exchange for educators
Simone explains how she focuses on the design of COIL/VE courses and assignments.
“One of the main advantages is that you can use ChatGPT to come up with ideas for how you can create and design a COIL course. It’s especially helpful in creating something that’s really interdependent, so the students rely on each other.”
But she says it’s useful to take it beyond the design to the assessment phase of a collaboration. She is interested in how to design a COIL/VE course using AI in such a way that it leads to collaboration between students.
Her concern, however, is that the human ‘messiness’ during interactions isn’t lost and that AI is used in such a way as to enhance this.
“One course that I taught was about human movement technology. Students had to learn a new physical movement. So, one group had to teach the other group a sports movement and the other ones had to reenact that movement and receive feedback on it. Of course, you cannot use AI to do this as it’s in real time. But the collaboration and the interaction remained. And they could use AI to assist them in giving the feedback in some way.”
Simone says this approach can be applied to other disciplines too.
The importance of preparing the students for collaboration during VEs
Mark says that during past training sessions, both he and Simone have felt the need to emphasise how important it is for students to be prepared for collaboration during virtual exchanges.
“I think because academics may well be experts in their subject area, they can feel less confident around how to prepare students for collaboration during a virtual exchange project. And actually, that is something that AI can help with. It can generate discussion prompts and a reflective activity. Or, it can make suggestions for reflective activities around potential collaboration. Those are very practical ways that AI can help with the design of COIL, without taking away from the power of the communication that happens between the participants.”
The concerning aspect of AI’s potential impact on COIL/VE
Mark and Simone highlight the idea that if we accept that we all learn through cognitive struggle and how enriching it is when you have a challenge, the problem with AI is simply that it often allows us to skip those challenge steps and get straight to the solution.
Mark says: “I think every educator out there will appreciate at the moment, and empathise with this experience of setting your students on a task. And some of them are just instantly putting it into AI, and then starting to copy out what the response is.
And it doesn’t feel as if there’s any thinking happening. So, it’s almost like we’ve gone back to rote learning. And that is a real risk, because without that pause, and that pre-thinking time, you are missing out on the kind of really cogitating about something.”
In fact, this is something that is already happening in the classroom all the time. Therefore, it will be happening during virtual exchanges as well. As soon as a topic is set by the tutor, it is certain that most students will begin with the AI response first, without doing any kind of reading of academic articles. Unfortunately, they’re just relying on this pre-processed copy.
The erosion of critical thinking
“There’s not much critical thinking going on there, which could be a risk’, adds Simone. “You can spot when something’s written by an AI straightaway. So, it takes away the genuine part and begs the question: how much of it is you? And how much thinking was actually going on? How genuine is this piece?”
Simone feels this erodes trust and is developing so fast and risks becoming overswhelming.
“The article is even a little bit tame in thinking about where we’re going with AI,‘ she continues, ‘ In the future, it will get so good that we will be able to communicate with a chat box or even a video that will appear to be somebody from a different country. And it will be so good that it won’t be real, but it will be trained so well that it could be, we don’t even need to work with somebody in India or South Africa or Ireland, because we’ll have this visual, like simulation in front of us.”
Taking away the human element
The concept that AI could be so authentic as to remove the need for human interaction in virtual exchange is a chilling thought indeed.
Simone goes even as far as imagining that it could even be better than just talking to one person from Ireland or South Africa or wherever.
“That feeds into our fears of an AI takeover, doesn’t it?, interjects Mark.
“Suddenly we’re faced with these avatars everywhere replicating the human nature of discourse. But in reality, when we look at student and staff comments about what they’ve enjoyed about their international collaborations, it’s this idea of connection with other people, this real sense of humanity, and the randomness that comes out of these random sparks.”
Simone concurs: “Humans are much more random than that, and we throw really random things out, and that’s what makes life interesting. So I still feel fairly positive that the human aspects of COIL can’t be replicated just yet. Because in a way we’re unpredictable.
But could it get so good as to be able to replicate our unpredictability? Only time will tell.