I am the secretary for the New Hampshire Council of Teachers of English, and in December 2022, we had produced a spring conference and the call for proposals was out. And I said to Dennis, "This new AI ChatGPT thing had just been released in November. And I figured because we are workshop-style teachers and we write with our students and we use reflective rubrics that we could create a session called How to Bot-Proof Your Classroom." And I thought that in March we'd be standing up there telling everyone workshop-style classrooms will alleviate any dangers of students using generative AI to cheat. And pretty quickly realized that that was not going to happen.

As we were prepping for the session in early February, a student of mine who was writing poetry, I was workshopping with him about a poem he was writing about the 10 reasons he plays football. And his draft was written in his notebook and it was pretty solid. And the next day I asked them to hand in typed rough drafts so I could give them more feedback. And his was wildly different, and it was in four-line stanzas and it had a weird rhyme scheme and very weird structure. And it was clearly very different and probably not him.

So I asked him about it in class and he very excitedly told me his older sibling, a college-age student, had told him about ChatGPT and he had used it to write this draft. And I said to Dennis, "Oh, we have to rethink our March proposal." And that kind of kickstarted our journey.

Even before Kristina's student showed up to the table with a AI-generated piece, we had colleagues around us who were scrambling with stuff that was hitting the classroom pretty quickly from ChatGPT mainly because that's what was the big release in, what was that, November of 2022.

So I was watching people around me struggle with the same ideas and struggle in a way that I don't think looked very appealing. It wasn't just a quick conversation with a student and then everybody moves on with writing a new draft. It was administration involved, it was parents involved and not necessarily believing that their child did use AI. So there was a lot of stuff going on even before that first student of Kristina's that made me think, "Ooh, how are we going to deal with this?"

And then as we started digging into how we could deal with it, a lot of the stuff that was being put out there made sense maybe from a quick reaction. Let's lock the classrooms down and everyone will write with pen and paper. Let's dig deeper and ask students questions that the bot could never answer. Things like this. We both started realizing that those weren't actually going to be very good long-term solutions. And we also started to see that even what we thought was our solution, as Kristina was saying, wasn't really a solution.

So that three-month journey between the time we decided to present about that and then the presentation itself, we pretty much showed up in front of the group at NHCTE basically saying, "We thought we were going to be able to tell you how to bot-proof your classroom. You can't do that. Here's what we're thinking." And then that really kind of developed this last couple of years where we've been thinking about how to actually integrate AI rather than push it away.

Yeah. Yeah, that's a big mindset shift. Do you think you're seeing a lot of that mindset shift as you're doing this work?

That's a tough one, huh, Kristina? Because I feel like we've both seen people really locked down and say, "It's not in my classroom." They also really want to or hope to use an AI identifier that will solve all their problems. So there's that camp, but we've also done two presentations now with NHCTE, and I'm thinking about the second one. People had a little bit more information, they had a little bit more thoughts and feelings about AI. And while we were trying to present about essentially our part two, here's how we've started working with it, there was a lot of pushback in the conversation that we were trying to engage everyone in. But I also saw people transform their thoughts and ideas in the moment, people who started with, "I didn't want to use this or didn't think we should be using this, but now that I've listened to some of the ways in which you are implementing it, I think I can do that too."

Do you use it as a tool in your professional life?

Yeah, absolutely. I was just doing some new planning for a new course that I'm teaching alongside an art teacher at the high school, a children's elective, and he's doing children's illustration. And we definitely used it to help us organize our syllabus and even as a planning units and collaboration. And it gave us this cool idea of using some myths, origin myths and having the kids write their own myths and having the other kids illustrate it. So it's oftentimes a jumping off point, or sometimes it's just to kind of fill out this chart that we have to do to show that we're doing our summer curriculum work. So absolutely.

We're still discovering ways in which it can be used. I mean, this shouldn't be a surprise, I guess, but discovering ways that it can be used that we hadn't actually thought of yet. Last week we were teaching at UNH Writers Academy and an author came in to visit and he showed a bunch of images as a prompt. And we didn't get through all the images. So Kristina and I wanted to use them the next day. And we were thinking about... He just used the image as a prompt and said, "Write from this." And we thought, "Well, we want a few written prompts that we can give them for these images." And we uploaded the images to ChatGPT and said, "Give us five prompts based off of that." And it was processing accurately what was in the image, and then giving interesting and unique prompts to get our writers going.

So yeah, we definitely use it in our day-to-day lives. And I think a lot of times I've talked about that as being a switch that you have to flip individually. I know I've had to do it, and then the flip automatically switches off and I forget and I have to flip it back on. But you have to actually tell yourself, "I have ChatGPT. I have AI as an assistant if I remember to use it."

It can be used in so many different ways that I never even thought of, which is kind of what I was just alluding to there. But we have in class and in the book written about using it as an image generator for the purpose of discussion. Give me five different book covers, for example, for Lord of the Flies. And then in class we're talking about which one is best and why.

And then one other little interesting note too is how students use it. So since we started this journey, it has moved away from just being a ChatGPT world. There are other programs out there, particularly for us, we were introduced to SchoolAI and Brisk Teaching this past year, and watching how students who can use those because they have a data protection agreement when they're kind of given the opportunity to interact with AI and we can view how they are using it is really interesting.

One other thing I would add too in terms of how my teaching has changed with AI is I think it's easy to see how it integrates in the writing workshop in terms of writing a personal narrative or authoring an essay. But even in whole class analysis, we read Lord of the Flies, where we read Macbeth and students are writing a two-page analysis on a character or something that they've chosen to write about. And I'm seeing my focus shift not from how well the essay is crafted, but how much their thinking has been pushed or changed.

So I'm looking at students come to these new insights about Lady Macbeth or about Jack from Lord of the Flies, and I'm impressed by how they're articulating, that their thinking has changed. But I'm not grading the quality of their writing. I'm looking at how much their thinking has been pushed by the book, the class discussions, but also the integration of AI.

It does hone in on another component of teaching and education that I think is just as important as teaching those basic skills of how to craft a five-paragraph essay, for example. That's kind of a baseline, what did you learn in school? But this pushing your thinking beyond what the teacher has the ability to even push your thinking to.

Yeah. And so the way that everyone's thinking is being pushed. Is that coming out of these really concrete prompting skills that you're incorporating?

Yes, absolutely. For example, in Lord of the Flies, we had students as we always have just track quotes that stand out to them or things that are interesting, things that they notice, things that they're curious about. And then I ask them to just do a written response. But I have them first prompt ChatGPT, and I talk to them about what makes a good essay prompt but also what makes a good ChatGPT prompt. And in the book we're writing about a couple of students who asked ChatGPT to pretend that Simon was elected the leader on the island instead of Ralph and what would happen. And ChatGPT is able to make out quotes.

Wow. Oh my gosh.

Yeah. And then the student's like, "This is really cool," and able to write this essay on these fake quotes, but that are still appropriate and still accurate to what he thinks the book would be if a different character was a leader.

That is very cool.

And another student was writing about... Yeah. Another student had her whole opinion kind of impacted about Lady Macbeth and why she was the way she was because of a way she was able to dig into, it was a chain prompt in that case. But she'd prompted ChatGPT about Lady Macbeth, and it had said something about her feeling guilty, and she asked a follow-up question to the bot, and the bot gave more evidence. And she was like, "I hadn't really thought of it that way." And so whether the essays are well-written or well-crafted wasn't our focus.

Well, that that information would be analyzed, whatever it is that the bot is producing. For example, your Simon example, just because it says that this is what would happen if Simon were the leader on the island in Lord of the Flies does not necessarily mean that that's what would happen because it doesn't exist. So the students engaging in that analytical experience by saying what they agree with or disagree with, which is where it's pushing those thoughts.

So do you still have apprehensions or things you're thinking about right now that you're concerned about?

That's a good question. I mean, the ethical use of AI I think is still a bit of an apprehension, with ChatGPT because it doesn't have a data protection agreement signed with most states, maybe not even any, because it uses the information that the users put into it to train and to fine tune. And I didn't even realize that putting a student's written assignment into ChatGPT for feedback could be a breach of their personal identifiable information at the very beginning. And we didn't have a policy in place in our district at the beginning of last year. So we sent out a permission form to parents to inform them that we were going to ask their permission to use generative AI and teach kids how to use it ethically. And 99% of parents were very excited about it.

I'm with you on and just to add a little bit to that. In our experience at the end of this year, we found out that one of the programs we were using that we were told had a data protection agreement changed it, or at least it seemed that they changed it from, if you using it...

Or they might be.

If you're using it for free, you don't have it. And if your school is paying for it, you do have a data protection. So my apprehension is about companies, big companies getting into education.

It must be interesting to be writing a book on something that is so rapidly changing.

Yeah, definitely.

I think it's helped keep our finger on the pulse maybe more than we would have if we weren't writing about it. And I think it's also helped us learn from it in a different way by writing about it too. I've always found that about writing anyway. One of the central reasons I am an English teacher is because I love writing, and I've always used it to try to learn from my own writing or other people's writing. But the process of putting something down on the page and then grappling with what it is that you put down on the page is a learning process. And I think the two of us have learned quite a bit about generative AI and about writing.