As a former High School English teacher, my perspective about AI basically boils down to two ideas:
- I can see it being a tremendous asset in the classroom. I could have saved so many hours planning and replanning assignments (or more likely, creating new assignments for students who were absent or on home instruction) with some kind of AI tool.
- I can see it being a tremendous hurdle to overcome. Before I left the classroom in 2021, I was already dealing with a good amount of cell phone use and students using their computers for anything but google classroom while in my physical classroom. At that time, if students had an AI tool like what is currently available (and unlimited access to it with no training) it would have made that experience all the more difficult.
While I believe it would have been challenging, I truly see generative AI use in the classroom as more of a hurdle than a catastrophe. It’s not something to ignore or be so frustrated with that you refuse to engage with it in any way, but it is complicated. And it’s okay if you feel that way.
My focus for this blog is to look at generative AI more like a former teacher, and less like an instructional designer.
I was a high school English teacher for almost a decade, and in that amount of time I amassed a GREAT DEAL of materials. Some were passed down to me, some were created the night before school, some were purchased from Teachers Pay Teachers… all of it lived on a hard drive that I still carry around just in case I ever need a particular file. (Which, until this blog, was incredibly unlikely. I think the urge to hoard materials is just part of being a teacher.)
Knowing that I could always comb through what I had to try and make something work, or to potentially go into my files in the future and update a project made me feel like I was always prepared for anything. It created this environment where the person who held all the knowledge was the expert and that provided a lot of comfort. And if we’re being honest with ourselves about what generative AI can be and what it can do to our classrooms, a lot of our issues stem from losing that comfort.
So how can we navigate this discomfort, while not ignoring the fact that AI is a tool that can transform our classrooms? I’m going to use a real, former assignment of mine to demonstrate how I would integrate generative AI into it. If it sounds interesting to you, feel free to use it!
I used to do a lot of Socratic seminars in my classroom. I specifically liked doing three rings of the seminar. The inside ring of students held the main discussion. They prepared talking points ahead of time, had questions ready to ask, and came to the talk with evidence they would pull from the text and their research. The second ring would be responsible for taking notes about the inside discussion, and they would bring up any of those notes and their own questions when the circles switched. I eventually added a third outside circle, and that circle was responsible for backchanneling on a website about the discussion. They would essentially have a live chat where they talked about what was being said in the inner circles (and they could do it quickly because they were typing), and they would also be responsible for equating all of the examples to real life evidence or evidence from mainstream media. A transcript of this live chat was eventually sent to all students so they could use it for help on writing projects. For my students who didn’t care for public speaking, the second and third circles gave them ample opportunities to be a part of the discussion. For my students who loved leading the discussion, the inside circle gave them a chance to lead and purposefully invite students into the conversation by asking them questions. These kinds of discussions were often quiz grades. When students first began doing them in my classroom they blanched at the idea that the quiz wasn’t a traditional sit-down assessment, but they always walked away with way more knowledge and a better understanding of even the most difficult texts, and they were able to speak to one another like experts in their fields. For me, that was a better barometer of their understanding than a ten question quiz. Plus, students were always in a much better place to write an essay or long form response after one of these seminars; the inherent collaboration was both beneficial and an early peek at what life would be like in a collegiate, collaborative environment.
But the last time I ran a Socratic seminar like that was 2020 and a lot has changed since. So knowing AI tools exist now, what would I change?
I would invite AI into the circle with us.
While students were doing the prep work for their seminar, I would allow them to use Copilot to create the first five questions or talking points. Generative AI is a great place to start brainstorming, especially for some students who find it difficult to get started. As students used Copilot, they would be required to complete an organizer that included what their original prompt was, the questions AI generated for them, and an assessment of those questions. Their assessment would include:
- Did the AI responses contain any obvious bias? If so, explain what bias you noticed.
- Sometimes responses can sound convincing and appear correct but include wrong information. Can you see how this may have happened here? If you believe it’s correct, how can you prove it?
- How did the AI responses address all facets of your prompt?
- What impressed you about this response/what disappointed you?
- In what ways is this response different from what you would have come up with?
Allowing students to use generative AI is a slippery slope, right? You don’t want them to create a whole essay with it, and allowing them to upload it into AI to “check for grammar” could quickly become an entire rewrite. That being said, allowing them to use this tool to help them get started, giving them a place to bounce ideas around, is just like putting them into small groups and letting them talk out concepts. The added benefit of the analysis is that they are thinking about the responses and whether they are helpful or not, not just copying and pasting them into their own work. It’s the act of thinking and rewriting that is helpful here.
After they use Copilot to get them started, the next five questions or talking points would be created on notecards within the classroom. These next five questions would be meant to be totally original, and almost feel improvised compared to the AI generated ones. While inside the seminar, we’ll also take a moment to discuss if using AI to create the talking points saved any time during the preparation, or if they felt that their questions were better.
So my original assignment that I’ve been using and feel comfortable with basically remains in tact, but I’ve merged this new idea into it to allow students to scratch their generative AI itch. As the teacher, I know the amazing benefits the students receive from this kind of project so I want to keep it around. And the more work they can do with each other, the more conversations they can have about the text before they sit down and write the essay, and the more comfortable they are with it all, the less they might rely on outside assistance. It’s not a guarantee, but it puts all of us in a good position to do good, thorough work.
If you’re looking at your own pile of teaching materials and wondering how generative AI might be threaded into what you already have, feel free to reach out to me at dpm6213@psu.edu. If you like the Copilot + Socratic method, and want to try it in your own classroom but need help getting started, I’d be more than happy to help. You don’t have to abandon all the work you’ve previously created, and you definitely don’t have to do it alone.
