While I understand the motivation for the question and agree with Prof. Heinke in terms of the practical applications that can be realized, I think there is a false assumption at work here. I am a language instructor at a university in Japan - in a faculty of education where I am engaged in English teacher training. While I do encourage my students to explore and experiment with ICT and AI, one of the most important elements of my teaching is an emphasis that language is a human endeavor. I offer to them the idea that the language classroom can be an oasis of human interaction, away from the reach of technology. In that classroom, we can experience the variability of language as we produce it and we can experiment with the potential of language as we shape it to our needs. If we replace the human interaction of the classroom with the appeal of some uses of AI, we are undercutting the justification for language learning in the first place. I prefer to keep my classroom in the human domain.

Using artificial intelligence in education offers potential opportunities and advantages—but it also comes with challenges and potential drawbacks. The potential benefits range from teaching assistance to individualized learning, while also we should not forget the ethical considerations surrounding bias, misinformation, and the future of educators’ roles. The intersection of AI and education is a dynamic landscape where innovation meets apprehension.

Dear all, I'd like to contribute to the discussion here.

Anthony Rausch : I understand your concerns and fully concur with you on the necessity to frame language learning as primarily human-to-human interaction. However, I'm afraid AI is here and it is going nowhere. As such, I believe it is our duty to equip learners and those who might teach them with the skills to make the most of AI and other media. In Dutch, we have a nice word for that: mediawijsheid (you probably recognize the WISE in WIJS). You can translate it as "media literacy" and the term refers to the knowledge, skills and attitudes that enable people to take active part in a complexifying world where media (including AI) take center stage. For instance, it has become necessary to rethink writing assignments as most chatbots can produce high quality texts in a matter of seconds. It follows that the 5-paragraph essay is a pedagogical deadend, and teachers need to start thinking differently about how they approach writing tasks. I am a proponent of the process rather than product approach. AI-generated texts can be a good starting point to discuss fact-checking (as AI is known for making things up), post-editing (an AI-generated text usually requires thorough editing to make it more human-like), etc.

This depends on the teaching method you follow. In my teaching environment the traditional methods are much more preferable.

This is a pressing issue in EFL (in Japan), my teaching context, which I think is an important distinction to make. There is no shortage of access to authentic English texts in all media today, and many universities (such as my own) may offer plenty of opportunities for students to interact with interlocutors face to face. There are probably many other languages where that is not as possible or practical, which makes the Chatbot much more potentially valuable for learners of those languages, in my opinion.

Teachers use voice recognition and text-to-speech technology with chatbots to simulate conversations. You can ask it to give you an IELTS interview prompt and it will evaluate your performance (quite poorly, in my opinion). Students can ask it to brainstorm ideas for writing or to proofread their work. Instructors will also use generative AI to help create writing prompts and other assignments, as well as use generative AI images to create copyright-free (I assume?) visual aids.

That said, while I understand that the original poster is interested in learning about teaching applications of AI, I think that the ethical considerations of using these technologies have too quickly fallen from view. I think OpenAI unethically used human labor to train ChatGPT (c.f. https://time.com/6247678/openai-chatgpt-kenya-workers/), which is what we're mostly talking about here. Also, the issue of how the training data was used in the first place raises legal issues (the New York Times is currently suing OpenAI over this), although opinions may differ. Finally (for the moment), if you are teaching younger learners, you must be careful because the terms of service for something like ChatGPT restricts it to users of majority age.

I hope this gives you some food for thought. I am personally utterly uninterested in the AI uses I have seen and experimented with thus far, and despite what the zealots will tell you, I don't trust AI to write academic English at all.

I think that it is interesting how the original prompt:

[How do teachers / lecturers around the globe use AI tools in language courses?] has generated (perhaps in part due to my posting stressing a priority on human interaction) a discussion of the 'place' or the 'value' or the 'whatever prompt you prefer' of AI (rather than the use of AI). It seems there are multiple groups: (1) those who see AI as inevitable and therefore worthy of consideration as to how to use; (2) those who see AI as offering valuable contributions to education and therefore worthy of consideration in terms of how ... and how much to incorporate; and (3) those who question an open-ended and unquestioning acceptance of AI as a foregone conclusion and would like to prioritize in a manner of socialization (with the terminology and mindset of 'media literacy' in mind) that there are alternatives to a fully computer-mediated existence. Part of what we do is to create alternative futures ... if we unquestionably accept what is offered to us through the social supply chain, then we are simply content and experience suppliers, not educators.

So I'm curious as to why interactions with humans is superior to interaction with machines. If a robot is able to interact with me in a way that I find enlightening, interesting, instructive, I don't see how that's different from interacting with humans. I believe that ideas like "language is a human endeavor" that Rausch argues, are problematic. For one thing, it isn't a human endeavor--other animals appear to have language, such as whales. What about intelligent beings that might live on other worlds? And why can't language be an endeavor of machines? Machines can learn and they learn through language--Google has had AI reading everything published for a long time in order to learn. I don't see why language is in any way limited to human endeavors. And I think the idea that it is, could lead to easy failure to respect the rights of non-humans who have language, including those that might be machines.

Responding to Traphagen: please stay on topic. The thread is about how teachers use AI in language courses - not a debate about human-machine interaction.

Anthony Rausch : I'd like to add a category to your taxonomy - those who see AI as inevitable as well as potentially valuable, yet nonetheless remain aware of the numerous challenges inherent in its use. The operating word is "challenges", which I prefer to the more commonly used "risks", because of its negative undertones. I think the future of teaching is in the hands of those who will not shy away from the dauting, yet rewarding task of reimagining new, future-proof educational paradigms.

Conscientious use of MT by students is to be encouraged. But teachers need to provide guidance on how best to do this. One way is described in our to-be-published chapter on how a teacher did this in practice. The focus was on how reverse translation can be useful for learning by intermediate-level learners. In brief, teachers can: 1) Set a writing topic, 2) Have learners write an initial version in L2, 3) Type what they wanted to express into MT in their L1, 4) Compare the output from MT against what they wrote first, 5) Adjust their initial text, through moments of noticing, 6) Check learning gains through a post-test one month later, personalized for each individual.

The results were very positive. It shows the benefit of students using MT as an personalized teacher, and letting them notice the benefit of taking personal responsibility for their learning.

The integration of Artificial Intelligence in language instruction holds significant advantages, including increased time efficiency and enhanced creative teaching techniques. As Artificial Intelligence becomes increasingly prevalent in both academic and everyday contexts, it is prudent to embrace its educational benefits for the betterment of both scholars and learners.