English Essays with AI - A Teacher's Rant
And I also teach my students how to use AI in their essays.
I don't think penalizing students for using a free technological tool they all have access to is the way to go. Half the reason it's so rampantly used for homework in the first place is because students are overwhelmed with schoolwork and homework most of the time.
Would students still use it if homework was limited? Most likely, yes. After all, if they can cut down the hours they need to spend on schoolwork and spend more time on things they like, of course they would.
But I'm not here to debate the workload and stress students are under. I have neither the power, time, nor opportunity to tackle that issue.
Instead, I deal with the AI issue, which I do have some control over for now. For one, I don't penalise my students for it.
In fact, I don't really give my students much homework either (which I have actually gotten a lot of complaints about from the school and a lot of parents).
When I compulsorily have to give homework, it's something like, "These are the spelling mistakes you made, so write each of them correctly 5 times." Or, "You got this grammar answer wrong in this worksheet, write the whole sentence with the correct grammar 3 times." It's individualised like that, focusing more on repetition - the correct way - to hopefully overwrite the wrong patterns that they had learnt previously.
Or else, I tell them all to review what we did in class, to go back home and watch the videos and read the notes and examples I covered in class that day or that week - both to review, and to see if there's anything they discover that they didn't understand. (Because I know that after a full day of different subjects, all of them are not going to remember everything.)
So, for one, because of my lack of homework assignments, students don't get to use AI if I give them a writing assignment in class.
But, on the other hand, as soon as AI was something I knew they all had access to and was using, I actually started giving them writing homework once in a while. I was like, "OK, I want you all to use AI and write an essay on this topic, since I know you will anyway from the complaints of your other teachers. These are the parameters I want you to write under and this is the word count range. Read through what you get, modify as necessary, write it down, and bring it in tomorrow. But, make sure you read through it and research any bit you don't understand, any bit you don't think is quite right. Change it, delete it, or research it and modify it. And only submit your final draft, because I will question you on it, and will be grading you based on how much you learnt from it and can verify the information in it."
Well, that first class of this, I actually had to change my criteria. Because it turned out that half my students didn't know how to write AI prompts properly. And even the ones who had good essays, they went way above the word count, and/or did not present the parameters I had defined for the format. I think about 2 students from that batch got it all right. But then, they got tripped up at certain points in the essay that they didn't quite understand but didn't know how to verify, or couldn't be bothered to.
So, no grades for that class. (Class grades wouldn't affect their overall grade anyway. I gave class grades mainly so that my students would not find my grading confusing when they actually wrote tests and exams.)
Instead, they got a lesson on the limitations of this rather new AI technology, and all the stuff that could go wrong. I reminded them that this was an LLM, a language learning model, and what that meant: It might be much faster than humans, but it was still learning how to parse information, and not all of it was going to be correct. (I had to repeat this a couple of times during the year. My students were more shocked that it was possible for technology like this to make mistakes.)
Then I taught them how to write AI prompts, how to define the parameters of what they wanted. And then, I added the parameter of mentioning at least 3 citations and resource links for each essay. That way, when there were doubts about the contents, it could be checked.
If any links were broken or paid resources, that didn't count and they would have to redo it. The first was because they technically didn't exist. And the second was because neither I nor the students were gonna pay for those for a 200-word essay, and hence those were non-verifiable content links.
The students had to check all of this before submission if they didn't want to lose points. And, of course, they still had to explain their essay like a presentation or seminar in class, while answering any questions that might come up.
Once students started bringing their laptops to school and use the internet, I did this directly in class instead of assigning it as homework. And once these essays passed muster in their typed forms, they had to be copied down in their notebooks.
Frankly, I got students to do more valid, relevant essays than they had in previous years before AI became a thing. Because the time taken for it had reduced. The teaching techniques were modified, yes, but more or less with the same goals as before. The only main addition was the verification of information.
Soon enough, the students became quicker at it too. Most of them completed more in a class, over time, and they started spotting errors and contradictions in information without my help. They self-corrected without my interference.
I myself ran out of relevant questions for them to complete from past question papers and such, and started using AI to get more varied questions in similar styles to give them.
And while some may not like to hear this, the truth is that students, unintentionally, read a lot of varied topics this way and consume a lot more information than they would have otherwise - especially those who don't enjoy reading for leisure. These were concise, 200-word (or less) pieces on various topics. None of those essays gave expert-level information, of course. Too short for that. But it was new information, all the same.
After a few months of my students getting used to this, I also had monthly to bi-monthly tests, where laptops were put away, and they had writing tests with exam-based essay prompts. And all of them showed remarkable improvement from where they had started.
Did all of them become exemplary? Of course not. Some still struggled. Some were average. Some floundered with test-based nerves and needed more time or went blank for a while. Some had other issues hindering them such as dysgraphia, slow writing speed, and so many other limitations and stumbling blocks. The students were still students and varied and human.
But, integrating technology in class - this new AI technology that many teachers seem hell-bent on saying is just for "cheating" - only helped my students in the end. Now, they know how to generate AI prompts, and they learnt research skills and verification skills that they might not have bothered learning until later. And, taking away the taboo of using this technology not only made classes easier for students, but also opened the discussion about the ethical use of AI and understanding the limitations of the tool.
This is not to say that everything is peachy with AI and that everyone must use them everywhere, of course. Rather, it is just to say that, instead of telling students not to use it, maybe consider that it would be better to teach students (and people in general) how to use it both effectively and ethically?
prguitarman
kianamaiart
titsay
s-u-f-j-a-n-s-t-e-v-e-n-s
izzy806
andulkaphoto
gummygunk
dailybridgerton
