Using AI to Deepen Learning with Ronald Lethcoe and Dionna Faherty

Please note, this transcript is generated by AI and may include some errors. 

Spencer Payne: Okay, we're live again with another episode of front of the class real stories from real educators with a double educator special here today, a deep dive use case on AI with Ronald Lethcoe and Dionna Faherty. And we'll let you both introduce yourself. Ronald, can you introduce yourself first and then Deanna of, hey, where do you teach? What do you teach? What's your role in the educational world? What's your background? Any fun facts? How do you like to introduce yourself to other folks in the educational world?

Ronald Lethcoe: Cool, nice to be on the podcast with you here, Spencer. My name is Ronald Lethcoe. I'm a curriculum and instructional design specialist at Clover Park Technical College. I've been here for three years. Before that, I was teaching over in South Korea. I was over there for 12 years, and I really got to learn a lot about what it takes to make a good educator, what it takes to...make students interested in learning. over there I got my master's degree in learning design and technology through San Diego State's online program. I love education. I love seeing people learn. And I think AI is a cool tool that can help with that. And I'm joined by one of our faculty members here from Clover Park Technical College, Deanna Faraday.

Dionna Faherty: I'm Dionna Faherty. I have been teaching for a long time. I teach English 101. Currently I teach 101, 102. And I have taught in the past, I've taught higher level, but I really enjoy teaching introductory level classes because I like to introduce people to things. I like to get conversations started and get the ideas out there more than I like looking at the nitpicky nuance of things.

I would say, fun fact about me is I used to be a theater major. I started as a theater major in college and I took a class on Shakespeare and I decided I wanted to teach that instead. I wanted to teach literature rather than act. that's, well, it kind of made it so I'm a better teacher in my opinion, because I have all this theater background. So every class is like a performance.

Spencer Payne: And we'll get into a little bit of the performance of this special use case of AI that you all have gotten into. But first, before we get into the nuts and bolts, can you just give a little bit of a hook? And Deanna, I'd love to hear from you for this of what is this special use case of AI that we're talking about here today? What did you do recently? And maybe a very quick little, what was the reaction from the students? We'll get into all the nuts and bolts in a second, but just hook everybody right now. Like, what is this use case that you unleashed on your students? What was it?

Dionna Faherty: Okay, well, I'm going to give some credit to Ronald here because he talked to me about building GPTs, which is making a very specific area of chat GPT specifically, meaning you use a little bit of the AI that's only for one purpose and you build it specifically. So I use it by creating three specific GPTs for characters for a book that we are reading in class that we are actually dissecting in multiple ways. And I have ChatGPT has talked to the students like those characters. So I create an assignment where the students go in and they ask the character questions about not only what they're doing, what happened in the book, but also what they would think about modern technology because the book is set in 1960, 1765, I think specifically. And that's probably the most creative use I've used AI for.

Spencer Payne: So we created effectively three AI characters from this book, The Outsiders, where the students can interact with the characters from the book and ask questions. In theory, go study, go understand, why did you do this? Why did you do this? How would you think about this today? So you created three characters from the book that the students can interact with. And real quick, before we get again in a nuts and bolts, what was their reaction to this? Did they like it? Did they hate it?

Dionna Faherty: Yes.

Spencer Payne: Were they leery? Did they enjoy it? What kind of interaction and engagement did you see after unleashing this creation?

Dionna Faherty: Well, I have to say that you can't take the students as a whole, so you're always going to have a spectrum, right? But it was overwhelmingly popular. Like I recently just used it online for an online class for the first time and actually had students in the comments saying this was a great assignment. I really feel like I understand the issues better from talking to the characters because the characters aren't limited to the book. The way that the assignment is being used is so they can kind of walk in their shoes and see how modern technology might've changed the story.

Spencer Payne: Yep, perfect. Thank you so much. And Ronald, you're kind of in the more approval, like administrative, how do we use AI, what's appropriate, what's not side. And you share with me, you had kind of a green, yellow, red approach to where might AI be allowed, useful, et cetera. Can you share just a very quick high level of kind of why would something fall into a green, yellow, red spectrum? And why was this use case that y'all cooked up a perfect example in the green that was approved?

You you thought it would work, all that kind of good stuff. So can you share a little bit about, about your approach before we ever get into the use case in the class? How did this happen from an administrative and a policy standpoint?

Ronald Lethcoe: Sure, yeah, and I'll give you a little bit of background for how I created the AI usage tags. we got kind of into this AI thing back in early of 2023 when I actually first started working here. That's when AI started to really get in the zeitgeist and have people using it. I would say it was right around the time that ChatGPT 4.0 came out.

That was when it really took off and the capabilities were much better than anything we'd seen before. We went to us in the Teaching and Learning Center here at Clover Park. We went to some conferences about AI. I will give a shout out to Teach and Learn with AI. They do it over in Orlando, the University of Central Florida.

Dionna Faherty: It's University of Central Florida.

Ronald Lethcoe: Yeah, University of Central Florida puts on Teach and Learn with AI every year, and we've gone every year. that's where we, the use cases that, you know, we're talking about a use case now, that's where we kind of got the ideas and like inspiration from what all the instructors are doing all over the United States. so one of the things that I saw was, I saw this table that had five different levels of AI use. And I thought it was really great. And I was recommending it to the faculty at our college to use something as a resource. Like, OK, you can go through your syllabus and you can say, OK, this is the way that we're going to use AI in this course. They were using it at the syllabus level. But I felt like something was missing.

And it was actually when I went we were giving a keynote speech at one of the other colleges here in Washington and Diana was there and we were, I was like thinking about this idea, like, it's a good idea but it seems like, you know, there's some levels that just are unclear about how to use it and so I was like, I just wanna drill this down to like, it's very essential. Like green means go, that means you can use AI for this assignment. Yellow, that means like slow down  look at the individual instructions in the assignment for how you can use it. So you can use it this way, but you can't use it this way. And then there's the red, which is the stop sign shape of the red. Don't use it. It's not going to help your learning in this instance.

And instead of making it a syllabus level where it's like either you can't use AI in this course or you're free to use it in any way,  we drilled it down to the assignment level. so you have the three together, but then each of them I created little banners that you can put on the top of your assignments if you're making assignments in Canvas. can just copy and paste and put it at the top and say, okay, this is a green light. That means that AI will not interfere with your learning outcomes because the main goal of having these AI usage tags is to make it clear about what learning outcome you're trying to achieve with the assignment and how does AI fit into that. Because if y'all haven't noticed, education is changing. you, Deanna used to give essays and have them, here's the instructions for the essay, write the essay, come back in and we'll look at it.

Well now, you can just take those instructions and have ChatGPT write the entire essay for you. And that's not gonna help for learning. that's why, depending on the assignment, you'll have a red tag, would, AI is gonna bypass the exact skill that you're supposed to be assessing. And if the learning outcome requires a student to demonstrate a particular cognitive process, AI can't do the work for them. And so, yeah, go ahead.

Spencer Payne: And in this particular use case of the three characters from the book that the students can interact with and ask questions to, can you share a little bit more of what was the learning objective you're trying to achieve? What are the outcomes you're trying to achieve? To use some of the words that you used there that put this in the category of green. Can you share a little bit about what were you trying to accomplish and why did this particular use case fall in that green category?

Dionna Faherty: Okay. Oh, no. Do you want me to?

Ronald Lethcoe: but only a go ahead Dionna I think this is this a you 

Spencer Payne: whoever feels best in best position to answer why was this green and what were you trying to accomplish? I'm guessing there's probably a little bit of collaboration from both of you. So feel free, Dionna, to start.

Ronald Lethcoe: Dionna, you can start and then I'll chime in if I need to add anything.

Dionna Faherty: Okay, so I actually believed that the AI was going to help quite a bit with the concept of, I guess this is overused phrase, but critical thinking, right? I created my English class by using a pretty simplistic novel of The Outsiders, which is actually a young adult novel. I expected my students to be familiar with it, but I wanted the base story to be simplistic so I could connect it to bigger issues like...

poverty, like the parentification of siblings, like class division. I wanted to be able to hook a lot of these larger scholarly topics we were talking about back to a base story. The problem I ran into was my students were not connecting at the level that I wanted them to with this story. They weren't seeing themselves. They weren't able to walk in the shoes of these characters. And I'm trying to say things like, hey, we're going to talk about leadership styles and we're going to look at how the main character of Derry Curtis leads his family and they're saying, well, he's not leaving his family. He's just some dude who lives in the sixties, who roofs houses like this doesn't. So they weren't doing the thing I wanted them to do. And so, and they, and they firmly believed that if these characters just had cell phones, the whole world would be solved for them. And I'm trying to get them to deeper meaning. Right. And so I created the GPTs and I said, okay, you know, have a conversation.

Let's answer cell phones and do ask them how cell phones would have helped. And they have these conversations where they realized, Hey, it, it, first of all, it doesn't solve all their problems. creates some new problems for them. And also they struggle with similar aspects of belonging and class issues that we have now. The characters allowed them to go from. Pat answers when I would ask questions like, how do you think modern technology would have helped the story would have kept the characters from dying, would it have kept the family from fighting, that type of thing. I would always get, yeah, it never would have happened because they would have had cell phones and Pony Boy would have never run away and it never would have happened.

And then they talk to the characters and the characters are saying things like, well, I was so mad at Derry that night, I would have just tossed my cell phone, it wouldn't have mattered anyway. And so I'm getting answers then instead like, 911 systems would have been useful if only Johnny had access to an online counseling or online chat group, if there were only other outlets. And so they thought more detailed and more specific. And that's what I was going for. I wanted them really to think about the characters and Ronald can tell you why that would be in the green area of using AI.

Ronald Lethcoe: yeah, definitely. It's, I mean, green is you're allowed to use AI in any ethical way. In this case, it's not, you know, here's an assignment, I'm going to go find out a way to use it. The assignment is the AI. And this is kind of where I feel like the future is headed, where we're interacting with AI and using it as a way to deepen learning. like, you know, she's talking about a pedagogical practice of deepening the comprehension and their analytical skills because they're looking you know they're talking to this character And it's giving them. I never thought of it that way So yeah, I really like the the way that diana's is using this in our class, and I think it's definitely Some of the where we're headed in the future for education

Spencer Payne: I never thought about it. Yeah, exactly.

Perfect, thank you. And I'd like to get into a little bit, if I'm a teacher, I'm an educator out there thinking like, well, I kind of want to do that. Let's get in the weeds of how this was created first. So what was the specific tool that was used to create these characters? Was it ChatGPT? Was it a couple of different things? Like, what was the specific tool or tools that this was built in and that the students were interacting with these characters in?

Dionna Faherty: Well, I use ChatGPT and I created GPTs and Ronald, if you could talk a little bit about the basic concept of how you create a GPT and then I'll explain specific nuances I used for this particular creation, if you will. Does that work?

Ronald Lethcoe: Yeah, so GPTs are a little bit different than traditional chat GPT conversation. Usually when you're interacting with chat GPT, you're asking questions, you get some response, and it's pulling information from the large language model. When you make a GPT, it's something that I believe is only available for people who have a paid subscription to ChatGPT there's a bit of an accessibility issue with it.

But if you are able to get the ChatGPT+, then you get access to anybody else's GPTs that they've made, and then it also gives you an opportunity to create your own. And the process is pretty simple. You just go in there and tell it what you want, and then you can fine tune it. It creates a little icon for you, and then...what else is really nice is you can upload documents. I'm sure Deanna will talk about it, uploading the text from the novel, The Outsiders, so that it draws the reference from that. And so it's almost like a conversation chat GBT that you would have, but it's very focused and there's one task.

And you can instruct it how to ask. You can say, if the student says, something that is not related to this book or this character, then respond this way and steer the conversation back to interacting with this character. Like, what did you think about the Seahorse game last week? The response would be, I don't know what you're referencing to, but maybe we can talk about something else. so Deanna can probably go into more detail about how she specifically fine-tuned it, but that's how you would put a GPT together.

Spencer Payne: And Dionna, first, I'd also love to start with just saying, what's the source material for these characters? And there's some maybe some copyright issues, like, can you just upload the whole book? Are you pulling from other people's opinions on YouTube about these characters? Like, what is or are the source materials to kind of create this character when you're not allowed to just like upload the book and and kind of feed it in that way because of copyright issues? So what is the source material to create these characters?

Dionna Faherty: Yeah. No. Okay, so I am going to tell you that I have a request in the Random House to be able to upload the book. Like, I like, please can I upload this book to my GPT? Because my fantasy of this particular assignment and some other assignments would be to be able to turn off the web. One of the things you can do when you build your own GPT is you can turn off the web access and it only goes from whatever you've given it. Okay? I have not been able to do that because I don't have legal access to uploading this book. But...

The thing is the, the outsiders has been around for over 50 years. So that means the story is very well known and I didn't actually have to upload anything. I opened the GPT the way Ronald's talking about it. And the very first thing I did is I asked it, are you familiar with the story of the, outsiders by Se Hinton written in 1967? And it says, yes, I am familiar with it. And I, I, then I moved to like, say if I'm building the soda pop Curtis GPT, said, I would like you to respond to my students' questions in the persona of Soda Pop Curtis from Tulsa, Oklahoma in the 1960s. However, you and then I go on from there and then it becomes this conversation like Ronald was talking about, like you fine tune it. I get it to where I want it to be as far as this is what I want it to sound like. You have like a split screen. I talk with my hands because that's my thing, but you have like a split screen.

So over here I'm building it and over here it's giving me some examples and I will ask it a question that I think the students might ask.  If I like the tone that they're answering in for the character, say, yes, that's the tone I want. This is the knowledge I want you to have. Now, of the issues, Ronald and I have talked about this, trying to get ChatGPT not to do something is actually harder than getting it to do something. So one of the challenges I came upon with this GPT is I want it to look at the book and the book only. And this particular story has adaptations in...a a TV show, and a Broadway musical. And I tell it, you know, only the book. But I'm always going back to check it myself to see if it's decided to take a trip down musical fantasy land or something.

But usually, I've had pretty good luck so far. So I build it that way. Now, I know as far as source material, like in your pre-questions, you asked, like, I include interviews with the author? Did I look at...scholarly work on this, which of course there's a lot. However, I didn't because I want this character to answer from the inside. want it drawn from the inside. I don't want it created from outside perspectives of what it is. The only slight change I made is I did, I did, I don't know how to say like narrow its 1960s information because I want my students to be asking the characters about what sort of you know, what's going on in the 1960s, specifically technology, but other things. But if you were to go into chat GPT and ask it, Hey, tell me, you know, what is going, what would be going on for with a 14 year old boy in the 1960s in Tulsa, Oklahoma, it's going to start talking about the space race and Vietnam and all kinds of things that don't pertain to this particular situation.

So I curb it to by saying, these are the outcomes I'm working on in my class. We're talking about the overall impact of poverty. We're talking about class division. We're talking about leadership styles. And I keep the GPT in that area. And I do actually upload some information about leadership styles because I want it pulling the exact leadership styles that I want. Is that all the info?

Spencer Payne: And yeah, and well, and I'm curious, I'm doing a couple more things here. First of all, just macro. That sounds like that was a lot of iterations and a lot of time and a lot of you playing with it and like, nah, that doesn't quite sound right. No, take out the bot space race. Can you share a little bit of like, before you actually unleashed this on your class for the first time? I mean, was this like,  prep time 40, 100, like all summer long for six hours a day. Like, can you share a little bit about like, if someone's listening, like that sounds cool. I want to try that from a time and a amount of iteration standpoint, like what might they expect given your experience of kind of how long it took before you felt like, wow, this, this feels right. And I'm, ready to go kind of unleash this on my class.

Dionna Faherty: I have to say that I'm not really sure the amount of time, it was...

Okay, look, this is all gonna have to be edited out, this right here. I'm gonna say it's at least 10 hours for the first one. The good thing, but it's not like I sat down for 10 hours. I worked and then I left and then I came back. The other thing is they, some of it I could reuse. Okay, so for whatever reason, I started with the middle brother, which is Soda Pop Curtis, which is why I gave you that example. I put in everything I wanted him to know about modern technology and I...

Spencer Payne: Yeah.

Dionna Faherty: did all of that already. I didn't have to do as much trial and error with the other two characters. And one of the funny things, one of the pushbacks I got from my students is they wanted different characters. I only did the three Curtis brothers and partially because of time. So they had a lot of the same information. They had the same background story. And then I could make that work. But I did have a couple of students were like, well, I want to dally. my immediate reaction was no,

Spencer Payne: No, too much work.

Dionna Faherty: because no, so I think I will have to get a few more months away from it before I do it. I will say I loved it. I was having a lot of fun and Ronald can certainly attest to this too. Sometimes you get in the weeds and the AI and you're there for hours and you don't realize it. The other thing, and I'm probably going to ping this back to Ronald because I have not had as much success in editing the GPTs once I create them. So I wanted to make sure it was really solid once I hit the...

Because you do it all, you get it where you want it to be with your two windows, and then you push the Create button, and it's there. It's solid. When I have gone back into Edit, I have not had as much success. But that could be me. I do want to say one thing. My students can access my GPTs without having a ChatGPT membership. So I link.

Ronald Lethcoe: yeah, that's important to note because yeah, in order to create the GPT, you need to have a plus or pro account. But once it's created, you're able to share the link directly to your GPT. I think there's some limit to the amount of interactions that you can have with it. But people will be able to use it. Your students will be able to use it once it's created.

Spencer Payne: Got it, okay. Got it. So two questions and back to you, Ronald, start is so by doing it that way, then well, two questions, I'll let you answer in either order. Right. But Dionna mentioned about editing. Can you share what does she mean by editing? How did that work? How did you pull that off? And then two, how do you make sure that the students are going through and editing the work after you already get it to where you want it to be? And maybe that's handled and they don't have the pro subscription so they can't.

Maybe it's that simple, but I'd love to hear your response to those. how did you go about editing? What do you mean by editing after the characters are created? And then how do you make sure that the students aren't overriding whatever you did so they can kind of recreate the characters in their own world?

Ronald Lethcoe: Yeah. Yeah, so the editing works. You can think of it like this, where there's two parts to the GPT. There's the creation and the editing function, and then there's the use. know, Diana went over how she created hers. I went over the process of how to create one if you want to do it on your own. The editing, I don't know. For me, there's basically what happens, I make the GPT, I click the edit button, it takes you into this backstage view where you can see all of the instructions that the GPT has. And so what I like to do if there's something that is not working right, like for example, we created a GPT for the college, we've been working on this project to align our college wide learning outcomes to the program learning outcomes to the course learning outcomes all the way down to the assessments.

And the goal with this GPT would be to take your assessment as it is now, toss it into the GPT, and then it will go through and it'll tell you how it aligns to the college-wide level and gives you suggestions on what you can do to make it, you know, make your assignment more inclusive or make it more transparent. And so we added all these instructions into the GPT, but there were some, sometimes it would come up with a different result or maybe it would focus too much on this thing and not enough on this thing. And so you would just go in and say, make sure you focus on the specific alignment level when you give your result or make sure that the transparency in teaching and learning is focused and make sure that there is something, you know, more examples of how you can use UDL in your assignments. And so there's some tweaking that you can do.

There's one way that you can tweak it where you're like basically having a conversation and saying this, and then there's another way to tweak it where you're like actually going through and adding and editing the, essentially it's not code, but it would be like the lines of code in the background. So it's just the instructions, editing the instructions. And there was another question.

Spencer Payne: And yeah, how do you make sure that you're the only one that has edit access, so to speak, and your students don't?

Ronald Lethcoe: yeah, okay, so and for the editing, that is strictly for the person who created the GPT. The students will never have access to that level because the GPT is created on Dionna's account. so, Dionna can go in and change it and edit it, but the students will never be able to do that.

Spencer Payne: Got it. Okay. And now, now Diana, I'd love to put, let's put our, all of ourselves in the students shoes when this thing is unleashed. So you put in your 10, 12, four, whatever out a lot of hours to, to make sure this is right to create all three characters. Right.

Dionna Faherty: Well, that's it. To create all three of them. I did take them. And it included some research of me figuring out what was going on in the sixties in Oklahoma.

Spencer Payne: Yep. And as well as what to, what answers to pass on, like when did the Seattle Seahawks organization start? Was he watching football games? Yeah. Not right. that scoping and all that editing. now we're there. You feel good about these characters and the interactions you're having. Can you share a little bit about that first week from the student's perspective when they're interacting with this thing? what are they telling you? Are they excited? Are they like...Check out this weird answer. Are you able to see the questions that they're asking it and can and like what what were your expectations and what were the results based on students engagement and interactions when this thing was first unleashed?

Dionna Faherty: Well, my expectations were that they would think a little deeper. Like that was my expectations. Like think more about these characters as people. And it's a funny line to walk because I did say a lot of times, please remember that this is an AI created person based on a fake character. So you can really get pulled into talking like this is a person. I'm, and we've talked a lot about that in the TLC in our world about making it not a person. And here I go making these characters. But I actually the first time I did it, I did it in class. I did it in class while I was standing right there. And I actually put the students in groups. And I said, remember the assignment that we did, because we had done an assignment before where I had them talk about modern technology in the outsiders and how it would have changed the story. It did not go how I wanted. They were far too simplistic with their answers. And so then it led me to create, like I was saying, create this.

And I put people into groups just to make sure not everybody's talking to the same character. And I said, okay, ask them questions about these. I put scenes up on the board. I didn't want to direct them too much. And I said, ask your character about what the scene was like for them. And if they could think of a modern technology that would have helped, or you can ask them about a specific modern technology. You just can't say cell phones. I kept saying that I was like, you can talk about an app.

You can talk about something about cell phones, but you just can't say cell phones. And it was so much fun. I can't even tell you how much fun that class was, that first class. I had groups on one side that whatever reason, and you don't even think about what these students are going to ask these characters. One of them asked if Johnny had just been a part of a online gaming group, if that would have kept him from being so depressed because he was so unwelcome in his home, but if he had this online gaming group and it led them to ask Soda Pop what his favorite video game would be, because I have the characters program to know modern technology, because that's the whole point. And for whatever reason, the Soda AI answers for himself and everybody else.

He says, well, I like Mario Kart, but I think Pony would like Minecraft. And then I had a group on one side of the class go who's talking to the pony AI and somebody says, we're talking to a pony boy, ask him what his favorite video game is. And so they asked him and he said, he said Minecraft and they're like, oh yeah, so does this Minecraft too. And I didn't put that many parameters on what they could talk, like what modern technology they could talk about that first time. I did a little bit more later because I did have students ask if they would be on Tinder, which I didn't see coming.

Like sometimes you just don't see until you're in the weeds. I do have to say though, the answers were freaking hilarious to Tinder. yeah, and who would be good at it? Who would hate it? Like it was, and I did learn some things and one of the things I wanted to say about editing, back to what Ronald was saying about editing, I don't know, he's better at this type of stuff than I am as far as like getting into the nuances of that.

Spencer Payne: to whether they'd be on or not or what they'd...

Dionna Faherty: I'm having an issue with my Derry GPT where he just wants to ask the students questions. The other two don't. The other two GPTs do not do that. But Derry's always like, have you done your homework? Are you, why are you asking me about video games when you should be studying whatever? And I can't figure out why he keeps asking them questions, except I think that he might be saying too true to the character in the book. I have no idea. have to, it's something I'm to work on over the break. Stop asking the students questions about their study habits.

Spencer Payne: So you share a little bit about, know, we're asking some hilarious questions, some things that are engaging and fun and interesting. And you've also shared earlier on that, I mean, that wasn't your intent, but your fun is good. Let's have some fun to then open up the conversation to go to the deeper levels that you were hoping to. And it sounds like the students went there too, well, hey, I realized if you had modern forms of communication, wouldn't that have just solved all your problems?

Dionna Faherty: Mm hmm. Yeah.

Spencer Payne: Well, actually, no, because I may have been so frustrated, I wouldn't have been looking and I would have gone and done this anyway. it sounds like and correct me if I'm wrong here, you're seeing kind of a, I don't know, I'll use my own term here, like a surface level sugar rush fun, like, well, this is interesting. Let's ask it some wild, crazy stuff and see what happens. Right. But you're also seeing, you're also seeing a level of depth that you weren't seeing without these characters when it was just like...

What did you think about soda pop or what were the ethics of it? Or how would he interact with the world today? Like when they're interacting with these GPTs that you created, you're seeing the level of depth that you were always hoping to have that level of conversation with your students. And now you're actually having it. So it sounds like you're fair to say you're achieving both. Like you're having surface level fun and a level of depth that you were always hoping to have. Is that a fair assessment?

Dionna Faherty: Yes, and I would say that like I mentioned the online gaming thing because actually that led a student to start asking about online counseling. They were talking about being part of a group and then they were so I put them in groups around a computer. So one person they were talking about the online gaming and another student said, hey, if he could do online gaming, I bet he could access online counseling. Now I don't have a Johnny GPT so they can't talk to him, but they did talk to the other characters and say, what do how do you think life would have been better for Johnny if he had access to mental health counseling online. And that brought another student to talk to ask, well, what would it have been like if Derry could have gone to college online because he couldn't go because he had to do all these things. And the concepts of, I don't know, it's just like they were being a little arrogant before that thinking, oh, modern technology would solve all their problems. If they were just like us, like this is like, this wouldn't happen.

Then they talk to the characters and they say, well, I don't, even if I had this information, I don't know that I would have trusted it. you know, I want to have my, my, my boys or my guys talk about it, which led us to the concept of poverty or lower income communities really rely on community and on each other more than higher end communities. Do just a lot of the points I was trying to make really bubble to the surface after they had engaged and had a little bit of fun with it. But I do want to emphasize this...really worked really well when they were working together. Right. And I used it online, like I said, this last quarter, and I literally just corrected this assignment, I think, two nights ago.

But I wasn't able to see them working with it, but I had several students who said, this was a fabulous assignment. I really felt like I understood more about how poverty impacted this, or I really understood Derry as a leader better after talking to him, or anything like that. So it works online. I really, really like it when they are together using it, but that's my personal thing.

Spencer Payne: And you just shared a little bit about some of that feedback. That was exactly where I was going next. At the end of the semester or school year, whatever it may be that these students are kind of in this world, what's the snapshot of the feedback that you hear from them? Like not that first day they're interacting, but on the whole. You just shared a little bit, but could you share a little bit more of what are some of the things that the students are saying after having this experience?

Dionna Faherty: Well, and I do want to point out that before we get to this particular assignment, I've had them working with AI in much more practical ways or what would be considered practical is I consider this practical, but I've had them working with AI to help them summarize scholarly articles. I've had them using it to help them outline. I've had them use it to just ask questions. Like I actually often have it, have them interact with the AI to question its results, that type of thing. So.

I don't know that in the overall feedback, I received more feedback about AI use in general. Where I received the feedback about this particular assignment is usually in the moment or they put it in the comments in the assignment. And I have had several students do that. Say this is a great assignment, kudos to you. Not only did I enjoy this, I feel like I learned more. I had another student said I was really getting bored with the book until this point. Now I feel like I understand the connections we were making. So. I do receive that, but I think, like I said, in the end of quarter commas, I just get more about AI use because I use a lot of it.

Spencer Payne: Yep, got it. And I'll ask this of both of you to start with you, Ronald. you know, again, there's a lot of editing. There was a lot of guiding. There was a lot of stay in these lines, not in these lines. There was a lot of prep work that you all had to go do to kind of get the result you were trying to get. Can you share a little bit about like, you know, with this use case, the students are liking it...dare I ask about the ethics of the assignment of like, you how much do you trying to guide the conversation versus you want the students have a genuine interaction, you're hoping they get this resolved, but you also don't want to lead them and like have the chat have the GPT is kind of give them your thoughts on the answer, right? Like some of these are ethical problems where it's gray, it's hard, right?

So I guess, can you share a little bit about like, you know, from your perspective, when there's hard conversations in society, there's gray areas, like, how do you give the characters enough room to be gray without trying to drive the students to realize, well, that's the right answer, it should have been done that way. Does that question make sense? guess, like, how do you kind of think about the ethics of what you allow the characters to let in and what you allow them to not let in?

Ronald Lethcoe: think there's a lot of ethical concerns related to AI and whenever we create trainings about it at our college, we start with, you know, this is what is possible and, you know, this is kind of the outlook and where education is heading. But then we always follow it right after with the ethics. For this type of assignment, feel like, mean, Dionna can maybe speak more to it because she's closer to it. I don't really see any ethical problems with using the AI this way other than I have heard stories about people who have kind of like a special connection with AI. We're talking about online help that people can get through therapy and whatnot. And some people have attempted to use AI in that way.

And I feel like there still is a little bit of gray in how people interact with it. I don't know. I've read stories about people who have gone the full way of trusting and putting all their desires and hopes and dreams and then being steered in a negative direction.

I mean, I don't know if anybody went that deep with what Diana was working on. But that being said, maybe that inspired them to say, OK, if I can make this character, then I can go home and make my own character. And so I think letting them know, think coming back, like taking a step back and talking about AI literacy with them, one of the things that we're trying to do at our college is...should be ready maybe at the beginning of the new year is an AI literacy course for the students. We've got our faculty on board and they're going through the courses that we've created for them. they're learning about the ethical concerns and they're learning how to use it. But the student view is much more nuanced and you have to be a lot more careful and let them know that they're they're not actually interacting with a human right the honor was saying I felt like I was actually talking with this character but you're actually just interacting with an algorithm that is always going to try to give you the response that it thinks that you want.

So it's like a a little happy puppy that says you know there's always coming up to you and and and showing you love and giving you all the attention and you can tell it what to do and it'll go and it'll come back and say what else do want me to do? That, you know, at the surface level seems great, but there are some ethical implications that exist underneath where, okay, know, a hype man is great, a yes man is great, is what this AI is telling me, not the human, that I've, you know, attributed to being a human it acts like a human talks like a human Is this AI that I'm interacting with actually giving me the best advice and so a lot of people knock AI for critical thinking Yeah, you're not gonna be a critical thinker if you're taking an assignment and then having AI do it for you the looking at the response and wrestling with that and saying this doesn't seem right.

I'll give you an example. was using AI the other day because I was having some problems. How can I say this? on. I don't want to get too personal. I was having difficulties with a relationship that I'm in with somebody and the AI gave me this advice where I was like, no.

I would never, like why would you even suggest that? That's a terrible, like, and you know, it goes back to, it thinks that that's what I want to hear based on like the way that I'm interacting with it, but in reality, like if I, you know, take a step back and say, like, what? No, I remember like, like physically saying, no, that's not an option. So I think being careful and, using critical thinking, like putting on those goggles whenever you're getting any sort of response from AI is one of the most important skills you can have when using it.

Spencer Payne: discernment, still got to be discernment on the back end. As we do get closer to wrapping up, I'd love to just go through a couple of just kind of quick hitter rapid fire style questions. Since unleashing these three characters from the outsiders with the AI GPT versions, are there going to be more characters? Is there's another book you're going to tackle and or what if anything is next? What's on your radar?

Ronald Lethcoe: discernment. Exactly.

Dionna Faherty: So what I would really like to do is take a book that's in the public domain, like A Christmas Carol or like Pride and Prejudice and upload it into a GPT, turn off the access and just be able to work with that so I can fully control what the information that they're getting. I've had great luck with these three characters from The Outsiders because I was so specific about where to take the information and I mean, it's been great, but I constantly have to worry about stepping on copyright. I'm always saying, remember, these are characters written by this person at this time, and I don't want to overstep. So I worry about that. So I would love to be able to just upload a book and say, we're just going to work with this, just only this book, and see where that goes. So that's probably my next fun thing with AI.

Spencer Payne: And then Ronald, coming back to you from a course design kind of perspective, what would be your current hypothesis on where AI might have the highest opportunity for positive impact in education in the next one to three years, realizing that the pace of technology is changing quickly, but just based on what you're seeing so far, where do you see a great positive impact in the educational world from AI coming up soon?

Ronald Lethcoe: Sure, so the first thing that I think is going to be big is the personalized formative feedback at scale. So AI can give students instant feedback on drafts, practice exercises, iterative work. The more practice you have, the more feedback, more improved you will be and be more prepared for any assessments in the future. I think the interactive simulations and role playing like what Dionna has created here.

It makes these practice environments that you can't do what she's doing in the classroom any other way unless you hire actors to come in and act like those characters. Students are able to practice these skills, are able to test ideas, ask random questions about gaming and Tinder. It builds their confidence, and it's really fun and exciting for them based on what Dionna is saying. And then the last thing I'm thinking is differentiated instruction. I forgot to say up top, while I was in Korea, I also got my teaching certificate from Moreland University. And one of the things that I learned when I was going through that process was about differentiated learning. And it was something that I had been doing, but I never had a name for it, like leveling out...making different assignments for different students. I think AI can adapt to the complexity of the student. It can give alternate explanations that maybe this student can understand it, but this student's having a little bit more time, little bit more struggle with that. And so it can really meet the students where they are.

And I think that's huge for access and equity. And I think those three things, the personalized formative feedback, interactive simulations and role playing and differentiated instruction is where I think AI is really going to shine in the future.

Spencer Payne: And then one more same question for both of you. I'll start with Diana. Any other words of wisdom, advice, anything that you were hoping to share that we just didn't have a chance to or anything that you did share that it's so important you want to just kind of restate it one more time. So any final new or repeated words of wisdom on AI, GPTs, education in general out there to the teachers and prospective teachers out there who might be listening.

Dionna Faherty: The one thing I did want to emphasize or point out is that AI has such great potential to help students be better writers, to help them find their authentic voice, to encourage critical thinking, to do all of these things that people seem to be afraid that it's going to take away. If you learn to work with it in your classroom and we walk our students through, this is how you use AI to help you write a better draft, but keep your voice by doing these other tasks.

This is how you use AI to help you think more deeply or make connections between two things that seem not connected. This is what I believe our jobs are going to be. And I believe that it is our responsibility to start teaching students how to use it so that when they are out in the world, out in their job, whatever they do, they will be using it appropriately. Because otherwise they're going to treat it like Google, which it is not.

Ronald Lethcoe: Yeah.

Spencer Payne: And Ronald, same question to you. Any new or repeated final words of wisdom out there?

Ronald Lethcoe: Yeah, I'll just echo what Dionna said. think AI literacy is the first step, learning about what it is, how it works. There's plenty of trainings out there. I'll plug the CPTC TLC LibGuide page as a resource where we try to put as much information as we can to help people become literate and then hopefully fluent.

One of the things we're working on right now is applying AI. so using it, I would say the final word of wisdom, try to find some information about it, but then also just get in there and try it. Like imagine that you're talking to an intern that is very eager to do whatever you want. so have a conversation.

If you get the output and it's something that you don't like, just say, don't like this, do it again, do it better. The way that I learned it is by going in every day and testing its limits. This is a tool where you can really learn your own way. The way that you do that is just by getting out there and playing around with it.

Spencer Payne: Well, thank you so much. some very real, although some very AI real stories from real educators here today, unleashing some custom made GPTs of characters from the outsiders and just watching as students interact with these characters, both from surface level fun, engaging to the level of depth that you were always trying to have that conversation at that level of depth regarding that book. So thank you very much for sharing this use case and all the work that you put into it. Appreciate you.

Dionna Faherty: Thank you.