Subscribe
Share
Share
Embed
Kate Meyers and Nicole Davis, AKA Two Maine Teachers and authors the EduProtocol Field Guide: AI Literacy Edition, discuss AI, AI in schools, AI for teachers, and AI for literacy for students.
Kate Meyers and Nicole Davis, AKA Two Maine Teachers and authors the EduProtocol Field Guide: AI Literacy Edition, discuss AI, AI in schools, AI for teachers, and AI for literacy for students.
Descriptions of effective teaching often depict an idealized form of "perfect" instruction. Yet, pursuing perfection in teaching, which depends on children's behavior, is ultimately futile. To be effective, lessons and educators need to operate with about 75% efficiency. The remaining 25% can be impactful, but expecting it in every lesson, every day, is unrealistic. Perfection in teaching may be unattainable, but progress is not. Whether you are aiming for the 75% effectiveness mark or striving for continuous improvement, this podcast will guide you in that endeavor.
[00:00:03] Gene Tavernetti: Welcome to Better Teaching, Only Stuff That Works, a podcast for teachers, instructional coaches, administrators, and anyone else who supports teachers in the classroom. This show is a proud member of the BE Podcast Network shows that help you go beyond education. Find all our shows@bepodcastnetwork.com. I Am Gene Tavernetti the host for this podcast. And my goal for this episode, like all episodes, is that you laugh at least once and that you leave with an actionable idea for better teaching. A quick reminder, no cliches, no buzzwords. Only stuff that works.
[00:00:41] I'm very excited today to have two guests to talk about something that, uh, continually learning more and more about, and I'm very interested in, and I need to talk to people who actually know something about it, and that is about AI in education. And I have two guests today. Our first is Nichole Davis, and she's a PK through 12 STEM and computer science teacher at Vasselboro Community School.
[00:01:07] She is author, national speaker, and a Harvard Project Zero Digital Thriving Fellow with more than 20 years of experience in education. Prior to returning to the classroom, she served as the emerging technology specialist for the Maine Department of Education, where she led statewide professional learning and developed Maine's AI Guidance Toolkit.
[00:01:30] Nichole is the co-author of EduProtocols: AI Literacy Edition, and is passionate about helping educators thoughtfully integrate AI and emerging technologies to foster creativity, critical thinking, and student agency. Through her teaching, writing, and presentations, she empowers educators to create human-centered learning experiences that prepare students to thrive in an AI-powered world.
[00:01:53] She holds a master's degree in education and lives in Midcoast, Maine with her dog, Luna. Her co-author in the EduProtocols: AI Literacy Edition is Kate Meyer. Kate is a veteran high school English teacher and instructional innovation coach at Mount Desert Island High School in coastal Maine. Kate is a national board-certified educator with dual master's degrees in educational leadership and literacy.
[00:02:19] Kate is a fellow with Harvard's Project for Digital Thriving fellowship for the '26, '27 academic year, and as co-author of the EduProtocols Field Guide: AI Literacy Edition, she speaks nationally about the mindful and intentional use of AI in education. Beyond her school walls, Kate co-founded Two Maine Teachers, a consulting partnership with Nichole that delivers practical professional development in AI for educators.
[00:02:47] When I met these two, I learned a lot, and I'm thinking you are going to also. I think you're gonna like this one.
[00:02:54] Hi, Kate and Nicole. Welcome to Better Teaching: Only Stuff That Works.
[00:02:59] Kate Meyers: How
[00:03:01] Gene Tavernetti: are you? Thank you for having us. Oh, gosh. It, it, it's, it's a pleasure. And, talking about AI, it is like something new every day. You know, the last time I had somebody on to talk about, uh, EduProtocols and AI was Adam Molar, and he made a comment, like, "In three weeks, all of this will be different.
[00:03:20] You know, all of this is gonna change in three weeks." So, so hopefully you guys can catch us up a little bit, because we don't say, "I'm gonna look something up in the internet. I'm, I'm gonna Google it." I'm gon- You know, and it's kinda the way AI has, is now.
[00:03:33] Nobody knows what we're talking about, but we all, we all say it. So when somebody asks you about AI, either in general or in education, how do you guys talk about it? And whoever wants to, wants to go first.
[00:03:47] Kate Meyers: I think the simple answer about the AI that we're generally being asked about, which is generative AI, is that it is AI, it's technology that can generate an output.
[00:04:02] So it can recognize patterns, it can make predictions, and then it can generate language or it can generate images. that's the AI that everyone's kind of asking us about right now. Because, of course, AI is deeply embedded in, in almost everything that we do daily. so I think right now what people are really interested in is, is generative AI and, and what that means, going forward.
[00:04:30] Nicole Davis: I was gonna say, too, I think when, when you talk about it within the classroom or they know that you've written a book about it or whatever, they think you're all rainbows and butterflies. So I'd say the other piece of that is telling them that there's a balance with it, right?
[00:04:42] That it's not good or bad, that it's, we've gotta s- find somewhere in the middle.
[00:04:47] Gene Tavernetti: So, uh, generative AI, you described that. So what is another term to describe other uses or just so that people would have an idea what that would be called?
[00:04:59] Nicole Davis: Yeah, well, so it's important to know that AI has been for, around for a while.
[00:05:03] Um, and Kate and I both get when we do sessions like, "Well, I'm not using AI," right? And, like, if you, if you held your phone up in front of your face or you drove here today, that's also AI technology. Um, and it is all blending together, but, um, that predictive AI is the stuff that's been around. It's, it's, um, looking at existing data and making predictions from that versus, um, generating something new and looking at those patterns to generate something new.
[00:05:33] Gene Tavernetti: So, uh, talk about how, could each of you talk about how you got to become interested or involved with this aspect of, of, of AI and, and its, uh, impact that it has and will continue to have on education?
[00:05:50] Kate Meyers: Um, I was working for the State Department of Education, um, in a role that they have for educators called the Distinguished Educator.
[00:05:58] Um, and in our state, what our DOE does is they will hire educators out of the classroom, they'll buy out their contracts for two years. You get to work for the Department of Education on whatever the special project is that they have open, and then you go back to your classroom at the end of two years.
[00:06:15] So, I, I had just come out of the COVID year, I had just done hybrid teaching and virtual teaching, and I thought, "This is a good moment to step back and, and take stock of where I am in education right now." I took the distinguished educator position, and in my second year in that role, ChatGPT came out in November.
[00:06:34] Um, and it, it came out, um, with no warning, and it just became my job in the role that I was in to learn everything that I could about generative AI, and at the time, it was really just ChatGPT in that particular moment, and to talk with educators in the field about what that might mean for education. So I had a whole year where I was really just focused on learning generative AI, thinking about what that meant for education, and talking with educators about it.
[00:07:04] So I was really lucky, um, to be in the role that I was in in the time that I was in, um, because I, I just really got to take that deep, deep dive into what that meant. And then Nicole joined us.
[00:07:20] Nicole Davis: Yeah. So I also was a distinguished educator. Um, again, it was an opportunity to do, um, professional development.
[00:07:27] Um, and I lucked out 'cause, um, so it's, they do two years and, um, the t- the two cohorts overlap. So Kate and I overlapped for a year, and then because it was so much further on, um, I was like, "You know, keep me. I wanna stay in this role." Um, and it's funny 'cause at first, like, when I first asked, it was my second year, and it was probably November, and you know, you're already thinking about the end when you're headed back to the classroom and, and what that looks like, and I was like, "You know, you guys need, probably need someone."
[00:07:57] And they're like, "Nah." And then come February they were like, you know, "Do you wanna stay on as the emerging technology specialist?" Um, and absolutely, and so I got to go further. And I can actually, I, I think a magic moment for us, and I, I can picture it, was we did our state tech conference, um, ACTM, and we did a presentation, and it was, um, lots of people came, and, uh, I can remember sitting with Kate afterwards and saying like, "Let's do something with this."
[00:08:27] Like, "Let's write a book. Let's, let's do something." 'Cause we were both so excited about it.
[00:08:32] Gene Tavernetti: So you were so excited about it. You had worked together, this was several years ago, right? So you started doing presentations. So how is what you present on now different than what you were doing? Originally
[00:08:47] Kate Meyers: It's been such an int- I was just thinking about that today.
[00:08:50] It's been such an interesting ev- evolution really, because we started with our excitement about AI and its implications for the classroom. So we would often do,
[00:09:03] you know, um, tools and, you know, creative challenges and, you know, critical thinking challenges and, and, and it was really about using AI in the classroom. And our work has really shifted and, and I would say grown, um, into the, um, human framework for, um, mindful and intentional use of AI, which is, which is, uh, it's just such a big growth moment from being excited about AI and, and wanting to try things out, to really slowing down and thinking about how we use this tool intentionally and mindfully.
[00:09:41] So we've, we've grown up with the tool, I think, which has been an interesting, um, evolution, and a really exciting one too
[00:09:49] Gene Tavernetti: So you mentioned the, the HUMAN framework. What people don't know until they read your book, unless they've been in, in one of your trainings, is an acronym. Uh, what, what is the HUMAN framework as you describe it?
[00:10:03] Nicole Davis: So it is really a framework for intentional use and being mindful about your use. Again, back to that balance, right? Not going all the way, uh, to banning it or not going to using it all the time for everything. So HUMAN stands for H is halt, where you're really pausing and thinking about, "Should I use AI for this?"
[00:10:23] Um, uh, U is for utilize. My brain at 6:00. Yeah. Um, my brain is, um, kinda crazy right now.
[00:10:33] Kate Meyers: Yeah.
[00:10:33] Nicole Davis: Uh, U stands for utilize, so thinking about, um, like the prompting, right, and how we're using it, but not just the prompting, the tool. How are we utilizing this so that it's not replacing what we're doing? We're not cognitive offload.
[00:10:47] We are really thinking about intentional use. And then M stands for monitor, where we're thinking about monitoring for, um, accuracy and bias and, and how does that play a role, and how do we mitigate some of that? A stands for authenticate, where we're thinking about authentic uses and really giving clear expectations around that, and what does it look like in the classroom?
[00:11:10] And then the last piece is, is really the reflective piece, note. Um, thinking about how do, how have I used this? How do I wanna use it again? Do I wanna use it the same way? Do I need to disclose that I've used it? So again, just overall that intentional use, a process in order to do that. And you know, you can, you can give kids a list to say like, "You can use it for this, but not this, and this, and not this."
[00:11:34] But what happens when they get to a point where something's not on that list? We've gotta give them some judgment pieces so that they know any use. Should I use it? Should I not?
[00:11:47] Gene Tavernetti: You know, it's interesting, as a person who got involved with, uh, tech after I met Corippo, after I met John Corippo, and I went to one of his trainings and, um, I had never even designed a slide, okay?
[00:12:04] Because the, I, I had a guy. Well, actually it was a, it was a woman, but I had somebody do it. I ne- I had never done it. And I sat next to somebody, and a very nice young teacher, and she was ... Until I just stopped asking her 'cause I realized how much I was bothering her and interrupting her. And then I worked on it.
[00:12:26] Oh, I, I'm pretty good at this now. And then I went to the next training, and the next training, uh, said, "Well, we're not using that platform anymore. We're gonna use this one, and we're gonna..." You know, and, and not being a tech person, and I always know who the tech people are in the training because they have all their stickers on their, on their computer, you know, and I wasn't one of them.
[00:12:48] So how do we get across to those folks? Or who, you know, because you don't wanna just be talking to the people with the stickers on the computer. So, so how do, how do we do that? How do you talk differently to them? How do you get them on board?
[00:13:04] Kate Meyers: I have a, a secret weapon, which is a very specific tool. Um, and I, I don't necessarily promote tools, but my secret weapon is, is a tool called Magic School, which is an, it's an AI platform that is built by educators for educators.
[00:13:21] So I think that's really important when we're looking at what tools are we going to use in education. I think that we need to look at tools that are made by educators for educators, because we have very specific use cases. And, and the people who made this particular tool made it with safety in mind, made it with environmental concerns in mind, made it with, um, you know, copyright issues in mind when it comes to image generation.
[00:13:50] They just have really worked hard to kind of do as much as they can right and ethically, um, with this tool, and, and then made it as easy to use as possible. So it's a, there's a very, you know, low s- ceiling entry point to this tool. It's very easy to use. It's, it's very intuitive. And so when I encounter educators, and I, I still do all the time, who have never used AI before, that's my secret weapon.
[00:14:20] I'm like, "Oh, well, let me show you this one platform." I don't start with, you know, if I'm working with people who haven't used AI, I don't start with ChatGPT or with Gemini or with Claude or, or, you know, any of those big large language models. I start with that, um, kind of secret weapon that, that was designed for, um, specific use cases
[00:14:41] Nicole Davis: I think making sure that we meet teachers where they're at, and Kate's describing those situations and those scenarios where they're at.
[00:14:48] And so making sure that we're not coming in and it's not like, "Okay, tech people." Like, you know, they're, not everyone's a tech person. Yeah. And so understanding that we've gotta have different entry points. And then along that same line, uh, and I, I'm gonna repeat myself again, I, I've gotten so much feedback about the balance of it.
[00:15:09] When you come in and it's like, "We're gonna use AI all the time, it's everything, da, da, da," like that's not what this is. It, it has to be a balance. And, um, Kate, when she uses it in her classroom, that's, she's able to say like, "I'm not using it 24/7," right? And so when, when teachers hear that, it's kinda like a sigh of relief of like, "Oh, I don't have to use it all the time.
[00:15:30] It's not everything all the time." And it, it changes their perspective. And, and you can actually watch, like when we do a session, you can kinda watch them come in like, uh, a little up in arms, and then they, they just relax as they go with, as they hear the human framework and understand that we're considering all these different pieces.
[00:15:49] Gene Tavernetti: Well, it is- And it's- Yeah, go ahead, Nick. Go ahead, Kate.
[00:15:52] Kate Meyers: I was gonna say, as much as I talk about AI education, when I really look at exactly what I'm doing in my own classroom, I, I'm using AI 4 or 5% of the time. The rest of the time we're not using AI. I was just talking with some colleagues and I said, "Oh yeah, the poetry unit.
[00:16:09] Yeah, we don't use AI at all in the poetry unit." So there's four weeks out of my 18 weeks where it doesn't, you know, we don't, it doesn't even cross our mind to use AI because we just have chosen not to use it in that unit, right? So, um, it's really thinking about that purposeful and intentional use. And it, it isn't just using it because, you know, it's new and novel
[00:16:31] Gene Tavernetti: So you mentioned y- so you mentioned, uh, your use, you know, maybe 5% in that you are
[00:16:38] you teach high school. Uh, Nicole, you work TK through whatever. Is there a difference in the, in the grade levels where we're introducing the AI use to teachers?
[00:16:51] Nicole Davis: Yeah, absolutely. So I think, you know, it's interesting 'cause it's a little controversy. Uh, do I use this with younger kids or not? And there's
[00:16:59] Gene Tavernetti: some pla- I'm sorry, I'm sorry, I'm sorry, Nicole.
[00:17:01] I have to disagree with you. I don't usually do this. There's not a little controversy. Okay, there's- There's a lot of controversy There
[00:17:06] Nicole Davis: is a lot of controversy. I agree. I agree. And so the question is, is, like, do you use it with young kids or not? And so there are a couple platforms that's ... Like, you always wanna check safety.
[00:17:17] That's, that's the first thing. We wanna make sure, you know, what is it collecting? Is it okay with all the laws that are out, FERPA, COPPA, all those things? Um, is it okay in our districts? That's number one. The second piece is really thinking about what is ... what's your intention? So if I'm thinking about my young kids, they've gotta be aware, right?
[00:17:41] And if we think about something as simple as Alexa, right, how many young kids, and I'm thinking young, young, think that that's a human? And that's a, that's a problem, right? So there has to be an awareness piece that, that we teach kids, and that doesn't necessary- necessarily have to be that they're using a tool, but using things like Teachable Machines or AI for Oceans, which is from, uh, they just changed from code.org to CodeAI, um, tools that are teaching them how, um, these things are learning, how they're set up on data so that they understand that something like Alexa is not AI.
[00:18:20] And I think you can easily adjust that human framework so that it is for younger grades, and maybe you're not just talking about AI, but you're halting before we are, um, using any type of technology. And should I, you know, watch a video for this, or should I, you know, read an article, or should I grab the iPad?
[00:18:39] And then figuring out, again, you utilizing it, monitoring. We still have to monitor for accuracy and biasy and all these things, right? Um, so using that human framework e- even in younger grades. Um, so the, the piece about using the tool, I think that's gonna be individualized and school-based. Um, I've seen some great uses.
[00:19:01] I've used it with second graders where, um, they read a book, and then they got to talk to the character and talk about comprehension. Like, you knew whether those kids knew that or not because they were quick to be like, "This character would not say that," or, "This character would," right? Mm-hmm. And, and then they had, and the engagement, right?
[00:19:20] It was Humphrey the hamster and that whole series, and they created these hamsters afterwards, and talk about engagement. It just was a really great lesson. So I think it, it comes back to that intentionality and really thinking about the purpose. And I think right now, the conversation around tech, we have to come back to what is the learning?
[00:19:41] And even with AI, we ha- what do we want kids to learn? And then can the tech enhance that?
[00:19:49] Gene Tavernetti: I wanna go back to, uh, something you say- said because, uh, I think you two have a lot of knowledge about this, and we're using, like, uh, uh, I think a multiple meaning word when you said safety. Because normally safety, you know, y- and, and AI, you know, you think of, you know, all the things we read about teenagers, and they're having these, you know, these digital friends, and
[00:20:13] But that's not what you're talking about. You were talking about more like collecting data. Is that what you were, how you were describing safety? Could you talk a little bit more about that? Because I don't- Yeah ... I don't know if anybody even thinks about that when they're using AI.
[00:20:28] Nicole Davis: Yeah, I think, uh, you know, I've done a lot of teaching trainings, and when I think about this and I think about the classroom, we get, what, five minutes at the beginning of the year, a little video or whatever about FERPA and COPPA, and then what happens is we find this cool new tech tool, right?
[00:20:42] And then our tech people tell us we can't use it. And you're like, "What? You're just making it harder for us." And it's, it's not because they don't want us to use it. It's because we have to follow certain rules so that whatever's being collected for students is safe. Um, and part of that also is, is not putting things into AI, so that personal identifiable information.
[00:21:06] And I think there's a lot of questions around data that we don't talk about, right? So when we think about data, our picture is included in data. And when you think about Facebook, how many people have gone ahead and made themselves cute little Disney princesses, right? And that's great. As an adult, if you wanna put your picture into AI, go for it.
[00:21:26] Make yourself a Disney princess. But we can't do that with our kids' pictures because that is a set of data, and it's gonna be training off of that. So those are all pieces that go into this, and understanding what you're putting in the machine is data, no matter whether it's a prompt, whether it's your picture, whether it is anything.
[00:21:47] Okay. So we wanna be very careful.
[00:21:49] Gene Tavernetti: I'm a little dense here. Yeah. What makes it unsafe? What makes it unsafe if you're doing something in the classroom, kids are doing something in the classroom on some platform?
[00:22:00] Kate Meyers: So I think about, I'll expand on what Nicole said because I also wanna add, um, voice.
[00:22:05] Nicole talked about it's not just our, you know, it's our prompt, it's our images, it's, it's our voice. I use my voice with my ChatGPT, but I had to stop and think about whether or not I was willing to give up that privacy because, um, what we put into the machine is potentially being stored in the machine.
[00:22:24] And so you have to think about all the data that we give out all the time that's all over the place in everybody's system, and how often those things get hacked. Is what we are putting into AI, if we're using ChatGPT or Claude or whatever it is that we're using, if one of those eventually gets hacked, if that information gets out, would you be okay with that information being out in the world?
[00:22:51] And if not, then you shouldn't be doing that with your large language model , whatever it is that you are choosing- to use. And, and we have to think about, um, that level of s- of safety because we, we don't know if this information is going to get out or not. It could.
[00:23:09] Gene Tavernetti: You know, it's interesting that you said, um, you know, they might be hacked, my inference is that you think otherwise it's okay that they have it, that you're trusting the, the people, you're trusting ChatGPT, you're trusting all of these large language models, uh, with it.
[00:23:25] No? Okay. people can't see you, but you're shaking your heads and laughing. Okay.
[00:23:30] Kate Meyers: Just- No. Yeah, I don't- Yeah ... with, with my own use of AI, I use it for professional use with no personally identifying information. It does know my first name, and I- it does know that I live in Maine, but it doesn't have any other information on me other than that.
[00:23:46] Um, and I'm, I'm very aware of what I'm putting into it, um, and, and what I'm not. And the reason that I decided to, to put my voice into it, that, that I, I personally was okay with that. But again, that's part of the human framework is, is understanding, um, the intentional and mindful use of AI so that you can make those decisions on your own.
[00:24:09] I know my voice has already been collected in other places al- already, so felt like that probably for me was, was okay safety-wise. Um, but that's, that's where we, that's why we have to talk to students and, and to educators about the mindful and, and intentional use of this tool, because our students have to...
[00:24:28] Our students don't know these things, right? They don't know that if they use their voice in ChatGPT that it's probably being recorded and stored somewhere, and eventually could that become a privacy issue. It might be. And so we need to teach our students and, and the people we work with that we have to think about these things all the time.
[00:24:45] Nicole Davis: Well, and I think it's, it's a risk analysis. So when I think about, like, say using my credit card, right, there might be a reputable store that I'm like, "Yep, good, cool, I'll use that," or online, you know, something. Whereas there's other places that, like, you buy something and you're like, "Eh, I have the chance of, of giving my information," and it's not.
[00:25:06] So you have to kinda analyze that risk, um, and make choices around that. And there are laws for our students that we have to make sure that we follow so that their data is protected.
[00:25:18] Gene Tavernetti: Okay. Um, wanted to talk a little bit about, uh, something that came up when I was in, in one of your sessions, and I think it was you, Kate, said, uh, very excitedly, as you as you do, talked about, well, I don't, I, you know, and you started talking about ChatGPT, and Claude, and Gemini.
[00:25:38] And I, you know, so, okay, I'm a novice, so what do I need to know, uh, you know, a novice educator, what do I need to know about the differences and why would I, you know, would I ask Claude, you know, "When do I use you? Am I gonna get a real answer?" Or-
[00:25:58] Kate Meyers: I- it's so interesting because they're all programmed, you know, ChatGPT, and Claude, and Gemini, and Grok, and, you know, Poe, and all the others that are out there, they're- they're all programmed to behave, um, certain ways and they, they all behave a little bit differently.
[00:26:14] Um, and so I, I started with ChatGPT because that's what came out. Um, and so that's the one that I've really stuck with. Um, a- and I, um, I would say compared to most everybody else, I'm probably a heavy AI user. I have just started branching out into other large language models four years later, essentially.
[00:26:34] So I'm tr- I've tried Gemini a lot this past year because it became embedded in my, um, workspace through my district, so we ha- we have Gemini. And I, I like Gemini because it is embedded in my Google workspace, and so I can ask my chatbot questions about email, or questions about my calendar, or questions about my Google Classroom, and it's all integrated.
[00:26:58] So if I'm doing school-specific work, I might turn to that tool instead of ChatGPT. Um, I've just, just last week started trying Claude out a little more seriously than I have in the past. Um, and I thought the output was, was quite nice. Um, it just had a different tone to it, and so I thought, "Oh, so for some, uh, some different tasks I might turn to Claude once in a while."
[00:27:25] I would say that, uh, that's not the usual use. That's not how most educators are working. They, they're not, um, going down that rabbit hole like I do. And so I would say pick one, and try it out, and stick with it for a little while, and figure out what you like and don't like about it. And then when you start to feel really confident, try another one.
[00:27:46] And what I always suggest to people who are ready to try another one to do is take a prompt and put it in the one that you've been using that you know you like, and then take the same prompt and put it in one that you're new to, and see what the difference in the output is. Compare them, because they do have different training on the back end, and you might find that you prefer one over the other, and you might go, "This is too overwhelming.
[00:28:10] No, I'm sticking with one tool," and that's fine, too. Um, that's my recommendation.
[00:28:16] Nicole Davis: And I would come back, too, to that, um, like tech use in your schools, that you're gonna-- Your schools are gonna have different tech uses based on like, you know, their, their rules and things like that. Every school's a little bit different.
[00:28:29] So asking, "What can I use with my school computer, with my email address?" You can, you can choose, like you can do Kate's process at home, right, with your email. But also knowing like what can you and can't you put in there that's from school. So being careful of those pieces are gonna be important as well.
[00:28:46] Kate Meyers: And I think that it's important to, to note too, I, I don't know, a few years ago there were a thousand tools for, you know, for educators, and, and as the years have gone by, they've, you know, they've gone by the wayside or they've gotten swallowed up by bigger ones. And, and I would say there are fewer that, that the tech people, Gene, like you mentioned, that the tech people focus on at this point.
[00:29:07] I would say there's probably a handful that kind of keep coming up in places that I, that I am, in spaces that I am in where we are having these conversations. So that's good news because it, it, it takes some of that load off of educators to feel like they have to learn a thousand tools, because it was very overwhelming when all of those tools started appearing on the scene, um, for educators, and so now it's, it's less overwhelming.
[00:29:32] And that being said, because I'm a big proponent of find a tool that works for you and stick with it for a little bit, I'm also a big proponent of trying out a couple of tools that are maybe different from each other so that you can see the possibilities, because I think that's a sticking point, um, when I talk with educators who only see generative AI as a tool for cheating.
[00:29:58] It's because they haven't spent time with it to see and imagine all the other possibilities that there might be for this tool. And that's not to negate, you know, the cheating issue. Of course our students are using it. It reduces friction and makes cheating very fast. I used to have to walk to the local bookstore and spend 99 cents to buy CliffsNotes, and then I had to walk home, and then I had to read CliffsNotes, and then...
[00:30:28] Like, there was a lot of friction in my cheating process and- Yeah, yeah ... AI has taken that out. Um, and there are, I, it's just filled with possibility too, and the more time you spend experimenting with it as an educator, I think the more possibilities you s- you see in it.
[00:30:47] Nicole Davis: Yeah, we often tell kid- our teachers to break the tool, meaning go in there and act as a student.
[00:30:53] Kate says, "Channel your naughtiest student," right? And put in the things that they might put. Put in the, like, "I don't wanna work today," and see what happens, so that that way you can make that judgment call as an educator and see what they're gonna get, and also see, you know, what's popping up on my end as a teacher.
[00:31:11] Do I get to see, you know, things flagged if, if certain things are said or, you know, what happens, so you can really get an understanding of the tool.
[00:31:21] Gene Tavernetti: Well, two, two clarifying questions for me. The first one is, uh, you mentioned, you know, try different things and, you know, see what you like. Are there criteria that it, that you have for looking at it because you're a user, or is there general criteria you might share with folks to make some judgments about, uh, which they use?
[00:31:45] Kate Meyers: It's really task dependent, so it really depends on what I'm hoping to get from the output. For example, if I'm hoping to get something factual and I'm hoping to get, um, a source attached to an answer, I go to Perplexity AI because that tool is built the right- Is that the fourth
[00:32:03] Gene Tavernetti: one you've mentioned?
[00:32:04] That's the fourth one you've mentioned, Kate. I'm sorry. See what you're doing to me? Trust
[00:32:08] Kate Meyers: me, it's fun. Uh, but, but if I'm looking for factual research- Yeah ... that's where I go. Okay. And I still have to evaluate the sources, I still have to read the sources, I still have to do that work, but it- Yeah ... it does some of the, it helps with some of the, um, initial gathering of sources.
[00:32:24] If I'm doing creative work, I go to ChatGPT, although ma- Claude might become my go-to the more I, the more I use it. Again, if I'm doing, like, school-based Google Classroom, you know, email situations, then I'm in Gemini because all of that is attached. So the more you play with them, the more you'll discover that you might like one for a particular task over another.
[00:32:52] And, and again, it's okay if, if you just stick with Gemini because that's what your school gave you, and that's what's in your school workplace. Gemini's fantastic. There's, you know, there's nothing wrong with it. It gives you great outputs. Um, so it just depends on how much time you want to spend in the exploration of it.
[00:33:07] Gene Tavernetti: De- it's time you want to spend saving time.
[00:33:10] Kate Meyers: That's right, yeah.
[00:33:11] Gene Tavernetti: Yeah, yeah. So, um, is there a place to go ... So, uh, again, not to give you a hard time, but so we've mentioned four different, different spaces, platforms, whatever you wanna call them. What should I call them? Platforms? Uh- I call
[00:33:26] Kate Meyers: them large language models,
[00:33:27] Gene Tavernetti: platforms, whatever.
[00:33:28] Okay. Okay. Yep. Okay. Okay. So we talked about is there a place, can I trust going to one of them to tell me, you know, what they're best at? You know, and where sh- and where should I go? Yeah
[00:33:42] Kate Meyers: You can, but then you have to test it because- ... they're going to answer you, they're going to answer you, they're going to give you a positive answer no matter what.
[00:33:51] They're not going to say- Right ... "I'm really bad at such and such." Yeah. They'll tell you what they're good at, and you really do have to test it, but I have done that. Um, compare, I go to ChatGPT, compare what you can do to Gemini to Perplexity, put it in a chart, and let me see it. And then you can ask it questions.
[00:34:08] But again, you can't trust the answers, um, out of pocket. You still have to do, um, some exploration on your own.
[00:34:15] Nicole Davis: Well, and these tools are sycophantic so that what that means is that Yeah, they're gonna tell you what you wanna hear, and you have to know- Yeah ... that you're really confident about it.
[00:34:24] So, you know, everything I put into ChatGPT, it makes me feel real good. Nicole, that's a great idea, right? So we have to know that, and we have to... So we have in the book, too, we have a, a prompting framework that we use, and really you can use any f- prompting framework, but what that does is it gives you better outputs.
[00:34:41] So part of our prompting framework is really that test piece. So at the end saying, you know, uh, "Be critical of yourself and tell me something that's wrong with this output," and asking those questions back and forth so it's more of a collaborative process than just, like, one and done. And even then, it's not just I can use this output and go.
[00:35:04] I have to check things, and I have to use my best judgment and, and go through that process of vetting it.
[00:35:12] Gene Tavernetti: Okay. The o- the other question that I had, Kate, is that you mentioned in one of your answers you say, "I like Gemini. I like my chatbot." What's a chatbot? What does that, what does that mean?
[00:35:24] Kate Meyers: So, um, at the very beginning you asked, I think, what AI was.
[00:35:29] I think we started off by talking about generative AI that generates an output. The way that it generates an output is through the back end of it, which is called a large language model. The large language model is really the engine behind everything. So the large language model is what is working behind it to help make these predictions and understand patterns, and the way that we interact with a large language model is through a chatbot.
[00:36:01] So we humans are chatting with a chatbot that's calling upon this large language model in the background, and when it generates an output, that's what makes it generative AI. So our chatbots are ChatGPT, and Magic School has a chatbot called Raina, and there's Claude and, and, you know, all of those. So that's how we're, that's how we're interacting with the large language model that runs in the background.
[00:36:31] Gene Tavernetti: Okay. And so I'm trying to think back to the session that I was in with you, and you were talking about training your chatbot for specific activities. It's like it's stored. It's, it's like you could call on it to do a certain task. Am I right about that?
[00:36:47] Kate Meyers: Are you talking about when we, um, shared the gems that we had trained for tasks?
[00:36:52] I,
[00:36:52] Gene Tavernetti: I, I don't know. It was one of the, one of the four or five you, you mentioned that it was over my head, you know?
[00:36:59] Kate Meyers: Yeah, so Gemini has, um, little chatbots that you can program to do specific tasks. Um, so for example, in the workshops, we gave Gemini gems to the participants that were trained on helping you create that task that we worked on in the workshop.
[00:37:15] So we gave it information on the back end in that large language model, and then you interacted with it through the chatbot, which gave you the output. Um, and so lots of, lots of chatbots have those little features built in, so Gemini has gems and ChatGPT has GPTs. They all have their own little bots that, that you can make.
[00:37:39] The other piece that we haven't talked about yet, and this is becoming more prevalent, and I just tried my first one out this weekend, um, is agentic AI, um, which is, it's a whole new, um, world. And so this is where we are moving from generative AI that gives us an answer or can carry out a task to an agent that can actually plan, use tools, take multiple steps, check on what happened, and then continue working toward whatever goal you gave it, and it does that almost autonomously.
[00:38:18] So I'll give you an example. This weekend, I Google searched car services near me, and there's now an AI agent built into that search that I can click on. And so it asked me, "What's wrong with your car?" And I said, "Oh, it needs new brakes and rotors." Great. It's going to call local car places local to me, and it's going to ask for estimates to get my brake pads and rotors replaced.
[00:38:42] It wanted to know when I'm available, do I need this in the next 48 hours, this week, next week? I gave it a bunch of answers and then closed Google, and today I got an email that included information from five to eight local shops with the estimates. So what happened is the agent took all the information I gave it on Saturday, and then Monday morning when all these places opened, called them and spoke to them on my behalf and got estimates, and then collected those estimates and sent them to my email.
[00:39:17] And so now I have a list of body shops in the area and all of their estimates, um, for that car work. Did they,
[00:39:23] Gene Tavernetti: did they mask your gender so you didn't have- So you didn't have that problem?
[00:39:29] Kate Meyers: Yeah. I have no idea, um, how the chatbot sounded. It did say that it would use my name but not give any contact information.
[00:39:38] So it probably said something like, "I'm calling on the behalf of Kate Meyer to find out," da, da, da, da, da. And it looks like... I said to Nicole, "It'll be interesting to see how many, you know, local places hang up on this bot." Um, but it looks like everybody- Yeah ... everybody it called gave it an answer, so.
[00:39:53] Gene Tavernetti: Well, I'm sure a bot answered, right?
[00:39:55] Kate Meyers: Yeah. Yeah, yeah, it might have been. It might have been.
[00:39:58] Gene Tavernetti: I mean, you know, you know- Yeah ... um, I had this vision as you're talking about this. So you go to a, a large language model, and there are all of these little specialty things. It's almost like a, a coach saying, "Davis, you're in." You know, the, you know... And, um, and you're shaking your head, so it's kind of, kind of like that, a kind of a specialty thing, sending a left-hand reliever.
[00:40:21] Um, uh, so this is... Okay We've talked about some of the things that are, that are in your book, you know, but we haven't specifically talked, uh, about the book. And one of the things that, that I was so impressed with, you know, the, the sessions that I went to with y'all was the, was that you were teaching students how to do some of the things that you talked about.
[00:40:49] How do we evaluate what they're getting? And, um, never hear about that. We always hear about the cheating. So, so what are some of the things that, that you teach kids to do to e- to evaluate this tool, this AI, these AI tools?
[00:41:09] Nicole Davis: Well, and this is AI literacy, when we teach kids about how they work, um, how to evaluate outputs, and it's important, right?
[00:41:16] Because our kids are gonna do it anyways. They're gonna use this. They are using this. So let's take the time to teach them. One of our favorite EduProtocols is shoe bias, where we're really evaluating, um, bias in, um, images. Um, and you could do it in text too, but we've chosen to kind of start with images.
[00:41:35] We call it shoe bias it- because we start with shoes. Um, and we start with shoes because it's, you know, it's not heated or charged. You can look at, you know, three different pairs of shoes and say that the AIs have generated and say, "Oh, well, like, this is what's missing," and, um, you know, they didn't consider this perspective, and what about this?
[00:41:58] And the ki- the kids can do that, and I'd say the kids can do that at a very young age, right? Um, so they understand and can start to see that, like, when we generate images or generate text, that there's gonna be missi- missing perspectives. And then how do we evaluate and go further to say what perspective needs to be represented?
[00:42:18] So they're going in and they're developing or creating the images the, of, of a shoe that repr- they want represented, whether that is something that, um, you know, they relate to, whether it's something that's, um, you know, uh, we've seen some made-up shoes. That there's all kinds of things that could be part of this conversation, and I would argue that it's a conversation that we don't necessarily normally have in our classrooms, and there's lots of places where AI opens up those opportunities for those things.
[00:42:49] And, and this is something, you know, we start with shoes, but it could be that in science class, I always think about, um, I always use the example of photosynthesis as a science teacher. And when I think back to teaching about photosynthesis, I taught what was outside our window, right? Like that's, those are the leaves we're talking about, but are there other areas that are, you know, hap- other leaves or other plants that are happening with photosynthesis?
[00:43:16] And so those misconceptions can be pulled out, and you can have a conversation about AI bias
[00:43:23] Kate Meyers: One of the other things, um, one of the other workshops was on, uh, prompting, which is, is really important. You know, I talk to, to people all the time who are like, "Well, I tried AI, but I wasn't really impressed by it," and it's
[00:43:36] And I often say, "Well, what was your prompt? Like, what, what were you asking it to do?" And they said, "Oh, I, I asked it to write me a lesson plan on photosynthesis, and it did, but it wasn't very good." And I said, "Well, what did you do next?" And they were like, "Nothing. I didn't like it." Um, because people don't n- again, they don't know the possibilities around prompting, and I think that's where we get into this cheating conversation too, because we know, we, we think we know it as an answer machine.
[00:44:01] "Give me a lesson plan on photosynthesis. Give me the answer. Write my essay. Solve this problem for me." But if you really do prompting well, you're having it, h- you're thinking alongside it and having it kind of deepen everything for you. So you might say, "Here's what I wrote," or, "Here's my interpretation.
[00:44:21] What am I missing? What am I overlooking?" Or, "Here are three arguments that I have. Give me three counterarguments." That way you can think about it and strengthen your own work, right? "Ask me questions that would make my argument stronger." So we're moving from that idea of all it does is help you cheat, to it can actually help you identify misconceptions, missing pieces, um, things that you've overlooked in, in your own work.
[00:44:49] I'm working on a presentation right now, and I delivered the entire presentation, um, verbally to my chatbot, and I, and I said, you know, "Imagine you're somebody in the audience who feels a certain way. What would your response be to this presentation I just gave?" Because it can help me see, um, what I'm missing or what I have too much of.
[00:45:12] I- I'm not having it write my speech for me. I didn't say, "Write the speech." I said, "Here's my work. What, what am I missing? How can I improve this?" And th- and that thinking that goes into that, um, is much deeper than just prompting it to give me an answer.
[00:45:30] Gene Tavernetti: I mean, I was in that workshop, you know, with you when you were talking about prompting, and you, it's almost like you were a nag.
[00:45:38] You know, you were nagging it. It wasn't, you know, it's just like, "Okay, but what about this? What about this?" And I, and I think y- you know, that it's, it's, uh, you know, there's some, uh, analogous work for students and, and for the teachers, because the teachers can be just as lazy And I say that with all deference.
[00:45:58] I mean, they got a lot to do. Th- they don't g- they're not given a curriculum that, you know, and so they, they get this. I, I used something today, I was checking it out. Somebody had told me that this platform had improved quite a bit, and I was very disappointed at the, at the result today. You know, I would've, you know, parts of it, yeah, but other parts of it, absolutely not.
[00:46:22] You know? Um, so th- that was, um, and that's, that's in your book, right? How to do prompts. There are, you know ... How often do you ... Th- this is one of the, the questions that I had for, um, uh, you know, your Edu Protocol book. You have, uh, lessons in there that really teach about AI, like you say, AI literacy. So how often would you need to teach a lesson on bias, or would that be something that you might revisit based on content that, uh, students might not be that familiar with?
[00:46:59] Kate Meyers: Uh, that's one of the things that I think makes edger protocols work for this type of content is that it can be repeated, um, so that we can look at bias in a couple of different places. I have my students for 18 weeks. We're on a semester schedule. So, um, in a semester I look at bias with my students in AI outputs three times, and so it's just kind of three quick hits with the EDRA protocol on bias.
[00:47:27] And what I say to educators is, "You don't have to teach every EDRA protocol in that book. If you and your colleagues are all talking about AI literacy even just a little bit, and everybody mentions AI literacy two, or topics two or three times over the course of your semester, it builds. And students pick up on those connections and, and over the course of their education they will hear about AI bias, they will hear about great prompting t- tips, they will hear about different ways to reflect on their use," right?
[00:48:00] They'll hear those things and it, it will build. So again, I, you know, the, the, the feeling that people might have is like, "Oh, well Kate, you know, is, is talking about AI all the time in her English class," and I'm really not. I'm giving small hits where it makes sense in the, in the curriculum.
[00:48:17] Gene Tavernetti: Yeah.
[00:48:17] Nicole Davis: Well, and it creates a mindset.
[00:48:19] Kate, will you tell your story about the girls with the, and it's in the book, but the, the sparkles and things like that? 'Cause that's what we really want. We want them doing it on their own.
[00:48:29] Kate Meyers: Yeah.
[00:48:29] It was the first time I encountered AI bias in a, in a pretty obvious way. So my students were, um, making podcasts, which is a very English teachery assignment. Um, and they were trying to come up with a title for, you know, their group name, what they wanted their group name to be, and they were just spinning their wheels and taking way too much time coming up with this group name.
[00:48:49] And I said, "Why don't you turn to your chatbot and just brainstorm with that, see if it can give you a group name that you like?" So, you know, they get quiet and they get on the chatbot and they're, and they're brainstorming with the chatbot, and I just hear this eruption at the back of the classroom, and they're like, "This is garbage.
[00:49:04] This is crap. We can't believe it." And I'm like, "What is going on back there?" And they're like, "Every answer came out glitter, sparkle girl power," you know- "... pink puffiness." Yeah. And I was like, "Well, what did you say to it?" Like, what? 'Cause that didn't describe those particular students at all. And they said, "Well, we told it that we were four teenage girls making a podcast about the American dream."
[00:49:26] And I was like, "Well, wow." And my first response was to be like, "Oh, well what was your prompt?" But then I realized, I mean, it was the prompt, and what we were seeing was the bias that the machine had about who teenage girls are, right? Yeah. And, and it was just reflecting that back to them. Um, and so
[00:49:47] we just had a great, uh, you know, just a really organic conversation- again, that just fit in the curriculum in the moment, 10 minutes, and then that was that, and we all understood, and we moved forward with, with new understanding.
[00:49:59] Gene Tavernetti: Well, you know, another, another, uh, example of use of AI that I read about, that you r- you two wrote about was, you know, using, um... and then I'm not gonna name the platforms 'cause I can't remember them anyway. But, you know, where they give feedback to students on their writing. The, the, the teacher puts in a rubric, and, and I've heard teachers say, uh, I'm gonna...
[00:50:24] You'll find out my bias. I've heard teachers say, "Well, keep trying, keep trying until you get it," and the kids will... You know, the kids see it as a game to get to the, to the multiple points. And as a- I- if that was me, I'm thinking, okay, uh, what did they tell you? What did your chatbot tell you? Do you agree with it?
[00:50:44] Now make the change, and why do you agree with it? And it's gotta be something I just taught you or that you, you've learned how to write, and that's one of the things that you talk about in there, in there as well. Uh, so, um, uh- So we've talked about ... So that w- that was one thing that, uh, that you talked about.
[00:51:06] You talked about bias, you talked about, uh, the last thing you just talked about. What, what other things th- in, in with regards to AI literacy do students need to know about?
[00:51:18] Kate Meyers: Well, I think in relation to what you just said is, is what's missing for me in, in that story that you just told about writing feedback and, and AI, is that, that students are pausing and then reflecting and doing that metacognitive piece, and that's the, that's the final piece of the human framework for its n- its note.
[00:51:39] How do you note your use of it? How do you reflect on your use of it? So bec- so that you are stopping and doing that thinking about your thinking, as opposed to being focused on the potential gamification of, of writing, just trying to get those points or trying to- Yeah ... you know, trying to make the game piece move, right?
[00:51:59] Um, it's really, the learning comes in when you stop and you say, "What worked for me? What choices did I make that I am proud of?" I like to ask my students what they rejected and why they rejected it, because I think it's really important to teach them, because it doesn't ... It's not natural, which I think is really interesting to observe.
[00:52:22] But it's not natural for them to reject feedback from the chatbot, which I think is really interesting. My juniors have a hard, had a hard time before I taught them that it was okay to reject the feedback. I had a student say, "Oh, well, it told me I should do this, so I guess I should do that." And I was like, "Whoa, hold on."
[00:52:40] Right. You know, just because the chatbot said you should do that doesn't mean that you should do that. You are still in control. You are in the lead of this process, and I think that it is essential that we teach our students that they are in control of this tool. It's going to take slowing down and thinking about what's working and what's not working, and, and what they're listening to and, and what they're choosing not to.
[00:53:05] Nicole Davis: And I think this is another one of those opportunities where, you know, we don't necessarily teach kids how to take feedback. And I think even as an adult, I think about when we got notes back from our editors a- and we were going through and we're like, "Yeah, yeah." And then we had something that we're like, "Ugh," and it's like we're the writers so we can choose this, right?
[00:53:24] And, and kids don't necessarily have that, um, you know, agency always because, oh, the adult said it so I gotta do it. Well, let's teach them about feedback, and here's another great opportunity to do that through AI.
[00:53:39] Gene Tavernetti: It's, uh, I mean, there's just so much- To learn, and there'll be more to learn tomorrow. And, and I'll go to, you know
[00:53:47] Did you pass out stickers, by the way? I think you did, didn't you? You passed out stickers. We
[00:53:52] Kate Meyers: did,
[00:53:52] Gene Tavernetti: yes.
[00:53:53] Kate Meyers: Okay.
[00:53:53] Gene Tavernetti: Okay. So if a teacher just starting and they don't live in Maine, and they can't see you, other than buying your book, what would you recommend?
[00:54:02] Nicole Davis: So they can visit us. Uh, we have t- uh, a website, 2maineteachers.com.
[00:54:07] Um, and then we have a Facebook page as well, and you can get to all of our socials from there. We're on LinkedIn. Um, and we have, like, our shoe bias is there. Um, our human framework is in there. I think at the core of it, Kate and I are educators, and we want, we want students to be able to do these things.
[00:54:25] We wanna help teachers, and that's really at the core of things who we are and what we wanna do.
[00:54:32] Gene Tavernetti: Awesome. Wow. I could, like I say, I could spend all day with you guys. But, uh, do you have any questions for me?
[00:54:42] Nicole Davis: I have one for you.
[00:54:43] Gene Tavernetti: Okay.
[00:54:43] Nicole Davis: Uh, what's one thing that you've changed your mind over in the last year from the ed tech world?
[00:54:52] Gene Tavernetti: Um, I think one of the things that, that you guys have talked about, and that is it's not all or... You know, it's not all or nothing. You know, it, it can be, uh... You can find your spots. You know, even in the, uh, even in the EduProtocol world, there are people who are, you know, they are very little digital. You know, it's all paper, it's all paper and pencil, and you get the, um, you get the results.
[00:55:22] The thing that I haven't changed my mind about yet, and you guys didn't help me change, you just reinforced, and that is that y- you know, the, the teacher still has to do the thinking. I mean, again, it's not... It, it, it, it's interesting, you know, you know, you work TK, TK through 12, and you know that when you're working with your kindergarten teachers, they are so sweet, and they, you know.
[00:55:48] And then you work with the middle s- You know, the, our teachers, we're just like the kids. And so, uh, you know, they don't say they're cheating, but they're saying, "Oh, I'm having my chatbot, you know, doing this grading. I'm having... You know, they created that lesson plan for me." Uh, but they still have to be the boss, and, and that, and that's the part that, that worries me, uh, that you still, you know, whether or not you continue to write your own lessons, you need to be able to do it to be able to evaluate what it gave you, and that's my concern.
[00:56:21] You know, that, that, that, that's, that's my concern, and that hasn't changed. But, um, uh, and the fact that, you know, we're seeing a whiplash against tech in classrooms just proves that, you know, we can, we can take advantage of the tech, but it- kids don't have to be on devices all the time.
[00:56:42] Nicole Davis: Well, and I think that's exactly why we created the human framework, so that you're keeping the human in the loop always and in the lead.
[00:56:50] Gene Tavernetti: In the lead. The human in the lead. All right. Thank you so much. Any final words? Any words of wisdom before we go?
[00:57:03] Nicole Davis: I would argue try something. Like, just, just try something. You know, put a prompt in, see what happens. It doesn't have to be perfect, but just see what it does. And don't g- don't be quick to give up on it
[00:57:18] Gene Tavernetti: You know, I, I need some advice from you guys, you know? Alexa changed voices, right, lately, and, and, uh, she's way too friendly for me.
[00:57:31] Not that... You know, and she keeps asking me, "Who am I talking to?" And it's, and it's been a month. "What, what's your name?" And I haven't given it to her yet, and so So don't give it to her? Okay. All right.
[00:57:46] Nicole Davis: Make up a name.
[00:57:48] Gene Tavernetti: Yeah. I'll use my Starbucks name. Something, something that she can spell. Yeah. All right.
[00:57:55] Guys, it has been, it has been wonderful, and I hope I run into you again soon, and, uh, thank you for being on, uh, Better Teaching: Only Stuff That Works.
[00:58:04] Kate Meyers: Great. Thank you for having us. This has been great. Yeah,
[00:58:06] Gene Tavernetti: absolutely. Thank
[00:58:07] you.
[00:58:07] Gene Tavernetti: If you're enjoying these podcasts, tell a friend. Also, please leave a 5 star rating on Apple Podcasts, Spotify, or wherever you listen. You can follow me on BlueSky at gTabernetti, on Twitter, x at gTabernetti, and you can learn more about me and the work I do at my website, BlueSky. Tesscg. com, that's T E S S C G dot com, where you will also find information about ordering my books, Teach Fast, Focus Adaptable Structure Teaching, and Maximizing the Impact of Coaching Cycles.