It seems you can’t view a blog, newspaper, radio or current affairs show without coming across the topic of students cheating with AI. There are concerns about the students not learning anything and undermining their future education. However, even though students should definitely not be using AI, it can be a handy tool for teachers in the classroom.
Picture the scenario. You are a teacher in class, and your students are having a very productive day. So productive, in fact, that they have used up all of your lesson plan for the afternoon. You have to come up with a suitable ancillary topic quickly. But there is nothing suitable online. What do you do? You ask Generative A.I.
If you are unfamiliar with Generative A.I., yes, it is scary. Yes, it is new. But, no, it is not difficult to use. If your 13-year-old students can use it, then so can you. Basically, if you can text someone, then you can use AI. The technique for using AI is knowing how to talk to it. This is called using ‘prompts’. You have to be very specific in your language. Think of it as going on the internet shopping and typing in the word ‘sweater’. You will get thousands of different products. To find what you want, you must be specific, e.g. colour, brand, detailing, etc. It is the same with A.I. except, instead of product details, you have to be specific about what you want to generate; in other words, your prompt has to be precise.
One way to engineer an exact prompt is to use the R.A.S.E. method (as recommended by Oide). R.A.S.E. stands for Role, Ask, Specific, Experiment. This is called Prompt Engineering. First, select an A.I. or an L.L.M. (a Large Language Model). There are many different Generative A.I. Applications. ChatGPT, Microsoft Copilot, Google Gemini, Notebook LM, to name a few. The one you use depends on what subject you are teaching and what devices you have access to at school. The best way to choose one is to experiment with a few of them until you find one that is suitable for your needs. Then you can begin Prompt Engineering.
Role. As the name suggests, Role involves assigning a role and context to the A.I. This would include giving details of your role, the level of your class, and what you are trying to teach them. So, for example, an initial prompt could be: ‘I am an art history teacher explaining Modern Art to a third-year class who would know a little about it’ After looking at the results for this prompt, there will probably be a large volume of text, which is more than you asked for or really needed. So, you have to develop your prompt a bit more. This leads to the next step
Ask. You have to define the task clearly. This will narrow down the amount of data provided by the AI and make it more accessible. The following prompt could be: ‘Give examples of Modern Art’ This will get you a little closer to what you want. You will have some examples, but maybe too many. So, now you have to be specific.
Specific. You must hone in on details like reading age and the type of language you require. The following prompt could be: ‘Format and create a detailed explanation of Modern Art. Give three examples. The reading age should be 13, and it should be written in straightforward language. You have already combined the first two prompts so that you will be getting closer. But you may not have a suitable result yet. So now, combine all of the above and go to the next step.
Experiment. From Role, Ask and Specific, you have all the information you need. Now it is time to combine all the prompts and see what results you get. If the information is correct, then it can be used. However, you may need to tweak words and terms a little to get a perfectly suitable answer. Your final prompt should be: ‘I am an Art History teacher. Give me a set of examples of modern art for a group of third-year students. Provide a detailed explanation. Followed by three examples using clear and straightforward language. The explanation should be suitable for students with a reading age of 13. At this stage, you should have the outline and examples you need for a lesson plan.
The critical thing to remember is that Generative AI only knows what you tell it. If you don’t tell it you are a teacher, it will not know, likewise, for all the classroom makeup and level of material required. As a final point, make sure you revise all of the information, and then you can edit and save it. You don’t want to give a normal-looking exercise to students only to discover that there is something unsuitable in the middle of it.
Generative AI applications can be a valuable tool for teachers. It can generate timely information sheets or exercises for pupils of any ability level. However, to ensure that you are getting the best and most suitable results, the AI must receive suitable prompts. However, using Prompt Engineering processes like RASE, this can be easily achieved.
