If you can't beat them, join them: Include ChatGPT in your teaching

In the article "Prevent ChatGPT cheating: 10 tips for assignment types" we describe how to set written assignments to make it harder for students to use ChatGPT to create answers.

We have also explored the possibilities of catching possible cheating using AI detectors and plagiarism tools in the article "Plagiarism and ChatGPT – How to detect the use of AI in written assignments?". Experience in recent weeks has shown that it is very difficult to detect cheating (unless you know the student's writing style and can sense that something is wrong), and it is completely impossible to prove. There is no source or original text to prove students plagiarized.

This article's premise is that if we can't knock ChatGPT off the track, we might as well team up with it. Therefore, as educators, we must accept that reality contains an artificial intelligence that can produce texts. Once this premise is in place, we must join our students in investigating, demystifying, deconstructing, and critiquing the textual artifacts ChatGPT and similar language models produce. Of course, this requires that time be set aside in the teaching to conduct some thorough rhetorical analyses of AI-produced texts and subsequently discuss what we come up with. The result should be that the students and the teacher find out where it makes sense to use the tools and, perhaps more importantly, where it makes no sense!

We do not have a ready-made recipe for how this should be done, but we suggest what to look at and how to approach such an exercise. It is probably most obvious in the language subjects, but the methods can easily be used in all subjects where written products are produced.

In addition, many of the exercises will apply to both primary and secondary school and the university level.

Suggestions for the inclusion of ChatGPT in writing:

1. Discuss what a language model is and what strengths and challenges it has.
2. Have students examine ChatGPT's knowledge of a topic that interests them.
3. Critical rhetorical/linguistic analysis of a ChatGPT-produced text.
4. Edit a text written by ChatGPT.
5. Comparison of a human-written text and an AI-written text based on the same task statement.
6. Make ChatGPT improve its writing by fine-tuning input.
7. Make an AI policy at school/in class/in class.