• Anthropomorphize: To attribute human characteristics, emotions, or behaviors to non-human entities, often animals or objects.
• Generative: Capable of producing or creating, often used in the context of AI models that can generate new content based on input.
• Pretrained: Refers to an AI model that has been trained on a dataset before being fine-tuned for a specific task. It already possesses general knowledge from the initial training.
• Hallucinate: In the context of AI, it refers to generating content that is not based on factual information or reality. The AI may produce imaginative or inaccurate responses.
• Perplexity: A measure used in natural language processing to evaluate the uncertainty or unpredictability of a language model. Lower perplexity values indicate better model performance.
• Sitemaps: Diagrams or outlines that display the structure and hierarchy of a website, providing a visual representation of its pages and their interconnections.
• Quizzing: The act of administering quizzes or tests to assess knowledge or understanding of a subject.
• Cubot: A term mentioned in the context of AI tutoring, likely referring to a tool or service related to educational assistance.
• Scalable: Capable of handling an increasing amount of work, data, or demand without compromising performance, often used in the context of technology or systems.
• Generative: Capable of producing or creating, often used in the context of AI creating content or features.
• Succinct: Expressed in a brief and clear manner without unnecessary details; concise.
• Quizzing: The act of asking questions to test someone’s knowledge or understanding.
• Parlor tricks: Simple and entertaining actions or performances, often used metaphorically to refer to basic or trivial uses of AI.
• Iterating: The process of repeating or revising a process or cycle with the aim of improvement or refinement.
• Degenerative: A possible typo; it could be “generative” instead, referring to the creation or production of content.
• Pedagogical: Related to the methods and principles of teaching, instructional techniques, or strategies.
• Deepfake: A technique that uses AI to create fake content, such as videos or audio recordings, that appear to be real and often involve impersonating someone else.
• Confidentiality: The state of keeping information private and restricted to authorized individuals or entities.
• Ethics: The moral principles and values governing an individual’s or group’s behavior, particularly in terms of what is considered right or wrong.

This is the ten minute teacher podcast with your host, Vicky Davis. Episode 823 Cool Cat Teachers AI Overview, what works in the classroom, what doesn’t, and ten ways I’m using AI now. This is a special episode where I’m going to talk about my last two semesters of using AI in the classroom, how I teach it to teachers, and so much more. Today’s sponsor is modern classrooms project. If you’re ready to take a free course on how to center students in your classroom with modern pedagogy, join modern classrooms at modern now. I’ll tell you more at the end of the show and why it’s so important in the age of AI to update what you’re doing in the classroom. So in this episode, I’m going to cover five different things. I’m going to give a brief explanation of how I teach AI both to students and to teachers. I’m going to talk a little bit about what I did in my early adoption stage of AI. I’m going to talk about what AI cannot do, and then I’m going to move into my fourth section so you can give you ten examples of how I am using AI right now. And this is both in the classroom, but also how I’m using it, a blog, podcast, and that sort of thing in appropriate ways that respect the human intelligence that we all need to bring to it. And number five, I’m going to conclude with some comments. Now, for those of you who are reading my blog or having my newsletter, you’ll know that I have been documenting all the different ways that I’ve been learning about AI. 80 days of AI and hi, of course, hi stands for human intelligence. And one thing I teach my students is that if you are using artificial intelligence in a knowledge area where you have no knowledge, you are not able to supervise AI, and it’s difficult to use it effectively when you have no knowledge base. So you’re going into an area where AI can’t be a lot of help for you because you can’t tell if it’s accurate or not. But we as teachers and content area experts, there are a lot of things that we can do and help our students use it appropriately in the classroom. So first of all, when I started explaining AI to teachers and the first session that I ran back last spring, March, April with my teachers, I used to just show AI, and I stopped doing that because people kept crying. And it’s very disturbing to be a presenter and see people crying. They’re not really crying necessarily because of you, but they are crying, because when people see particularly Chat GPT or some of these other tools for the first time, it can be a very emotional gut response to it. So now when I go into a school or go to an organization, I typically will send an explanation video beforehand that people can watch on their phone. Really short, twelve to 16 minutes video I have that I use, and we load it into like edpuzzle or pear deck or nearpod and verify that they’re using it. And it’s part of the professional development, and that does help a lot. The most recent training I did for a school up in Indianapolis, it was great that I sent it ahead of time because we could get past that emotional response and talk about the positives and the negatives and address some of the early questions. Now, when I talk about AI, of course, everybody wants to talk about Chat GPT. I talk about the G, the P and the t. So G stands for generative. And down here in south Georgia, where we’ve had tornadoes, many of us have generators. And while people put different things in generators, a lot of people put in gas in the generator, and then out of it comes electricity into the house. So with large language models or llms that we’re using with chat, JPT, bingchat, enterprise, Google, Bard, all these different things, they are generative. We put in a prompt and it outputs language for us. Now, one thing you can’t forget is that it’s really more a conversation than just a one time prompt. If Chat GPT, or any AI tool I’m using, I use big chat enterprise a lot, gives me something I don’t want. It is your responsibility to tell it that that was not what you wanted. So it is generative. But then we moved into pretrained. So what has the AI been pretrained on? Now, this is really important to understand because we have these new gpts that those of us who have Chat GPT plus, which, yes, does cost $20 a month. So granted, not everybody has that. We can make a custom pretrained GPT, and I’ve made some and trained them on certain things. So, for example, I’ve made a tool that my computer science students can use for AP computer science principles. And part of the pretraining data was the 200 and something page book that the college board created that goes along with that course. And then I have told the GPT to follow along with those standards when it’s mentoring or helping my students on content. So it makes it a little bit more accurate than just the regular Chat GPT. But the question you always have to ask yourself when you’re using AI is what data has it been pre trained on? Now, this is what’s so exciting about some of the private tools that are coming out. Bingchat Enterprise, for example, I do some work for Microsoft, so I do need to disclose that although they did not sponsor this episode or anything. So in Bingchat Enterprise, for example, that is internal to your organization. So none of that data is being sent to Microsoft. When you’re using Chat GPT, just the regular Chat GPT, that information is retained at least for 30 days. You can, when you do a custom GPT, kind of tweak those settings a little bit more. But again, it’s pretrained. And then the last piece is that it’s a transformer. Now, this is a piece that a lot of teachers really struggle with because way back in the day, in the, even before that, everybody had a world book encyclopedia. And so if Johnny came in at 10:00 at night after baseball, he’s exhausted and he wants to write his paper, he might pick up that world Book Encyclopedia and copy something on penguins out of the World Book encyclopedia. Well, when he turned that in to Miss Adams, Miss Adams looks at it and goes, Johnny didn’t write this. And what does she do? She opens up her copy of the world book encyclopedia and then she can see that Johnny basically copied it out of the encyclopedia. Okay, so t is for Transformer. So if I record this episode tonight, then I record this episode tomorrow. I may have the same content, but I’m going to have different words around that content. I’m going to talk differently about it. My speech will be transformed from one time to another. So sometimes when you ask me a question, I might answer it one way on one day and a slightly different way on another day. Although I have that base content knowledge, when AI gives an answer to a student, it is not like you can copy that and put that into Chat GPT and say, did you write this? It’s not really going into a database and it’s not going to be accurate and be able to tell you if it wrote it or not. So be really wary of that. I had a student way back when Chat GPT first came out. This was back in January, and my computer science students and I were experimenting with it. And he says, hey, ms. Davis, come over here. I just took an essay that I wrote in 9th grade, and he was a senior, wrote in 9th grade, and I fed it into chat GBT and I asked it if it wrote it, and it said yes. And then when I said no, you didn’t, it said, oh, I apologize, I didn’t. Okay, so you can’t really know what it wrote because it is a transformer. Now, I’m going to talk more about that in just a minute. But one thing I want you to remember, and this is a stat I’ve learned at ISTi this past summer, and I’ve cited a ton in my newsletter and on my blog, is that around 85 million jobs are Predicted to go away because of AI. But over 90 million new jobs of humans using AI are predicted to be created. So knowing how to use AI is kind of part of what we want to do. But we’re going to enter the second stage now and talk about the early adoption of AI and what I experienced and how my students experienced. What do we do when we first go in? We say we feed in a question and we look at the answer it gives us and we go, oh no, this is really writing a pretty solid answer. What do I do about this? Now teachers who are just still assigning the ten questions at the end of the chapter are going to have problems because kids are going to put those in and the teacher is not going to be paying attention enough necessarily to catch that. Students are just copying and pasting and putting those in there. So our pedagogy has got to change because it can do simple answers pretty, pretty well. And so at the beginning we’re like, oh, well, how do we catch know there is a great Facebook group? And I say, great, but I feel differently about it. Different days. It’s on Facebook and it’s about Chat GPT. And so much of the conversation is Gotcha conversation. How do I catch kids using Chat GPT? How do I catch kids using AI? So the problem with this whole gotcha catcha environment is the bias that’s inherent in it. So my student William that I had on the podcast who did a project about Chat GPT versus Miss Davis and was looking at the assignments that we had done and evaluating which ones the students did better and which ones they did better, said, well, you know, I think sometimes kids are using Chat GPT for the easy a and really was taken aback by that because that’s terrible that they can get an easy a and that we’re not assessing any other knowledge and we’re so product focused. So as my friend Alana Winnick says, we’ve really got to start focusing on the process, not just the product. Because the other thing I’ve seen is that there are some students I’ve actually caught just going for the ECC. They know if they turn In a product that is completely perfect and everything’s spelled correctly and everything is just perfect, seeming perfect. Of course, they don’t have the content knowledge to supervise, to know if it’s accurate or not, but it seems perfect to them, then they know that the teacher is going to look at that and say, well, you didn’t write this. You can have a senior, for example, who could say, write this paper on Macbeth on a 6th grade level and include some common misspellings. Okay, that is a thing. They know how to do that. So some kids will go for the easy c or the easy b. You can make things write like you write and at the level you write. Now, we need to be thinking about this, because what this means is that your best students, your students who are already making the a pluses on their papers, are going to be the students who you are not going to catch. Quote, catch. And your weaker students are going to be the ones that you immediately assume they are being dishonest. Now, this really hit home. I was working with some teachers, and I had talked about the importance of moving from Gotcha and changing the pedagogical approach in a school. And I had a teacher come up and she was in tears, and she said, that was me. And she was talking about something that happened 15 years before. And she was weeping, and she said, I had this class. I was really struggling. I needed to make a good grade, and I was really struggling in the class. So we had a final paper, and I decided to put everything into that paper. And I had been turning in work that was c work, and I knew I needed to do better. So I started writing it, and I drafted it three or four weeks ahead of time. I went to the writing lab. The writing lab person helped me. I revised it again. I went back to the writing lab, and I spent a good three to four weeks working on this paper. And when I turned it in, I knew it was amazing. Well, when the professor handed back the papers, the only thing at the top of my paper was see me. And so I went to the professor, and the professor said, I’m going to fail you. You didn’t write this? And she said, no, I did write it and explained the whole process and even brought in the tutor that had helped her write it. And the tutor came in and said, no, she wrote this. The professor on her paper wrote b minus just based on principle, even though it was an a paper said that the student deserved a b minus because just did not trust that the student had done it. She said it destroyed her. It made her not want to try. It made her not want to learn. It destroyed her love of the class, and in many ways, it destroyed her hope in her ability to even get out of that slot of the c slot. Now, seriously, do we want to be doing this to our students? We are destroying our relationships left and right by accusing the wrong kids and letting the wrong kids slide by. We’ve got to change our pedagogical approaches. So some things that I’m talking to schools about. One is we’re writing the essays in classroom. Our brilliant writing leader at our school, dawn, is totally revising how we’re teaching writing and we’re writing in the classroom, we’re outlining in the classroom. She has something called a Kanban board that she’s using to kind of track how kids are progressing in their paper. And it’s just a beautiful process. And even at the point where they do peer feedback, she actually will have them go into Chat GPT, feed in the rubric, feed in their current paper and say, I need you to give me feedback. This is my teacher’s rubric. This is what I’ve written. Are there any places where this paper needs to be polished but do not rewrite it for me and give them points, bullet points on how to improve that paper is kind of like a first pass along with peer revision. And then you can hyperlink to AI chats. So they have to go in. If they’re using Chat GPT or if they’re using perplexity or whatever they’re using, they have to turn in the hyperlink as a citation. So that’s one big practice that I’m doing is my students are turning in hyperlinks as citations. Anytime they have a chat with AI about a project that they work on for me, they have to turn in that citation. So the other thing we have to remember is that another AI tool, photomath, has been around for a very long time. Many believe that this is why the math scores are plummeting on standardized testing, but that kids are still doing pretty okay in the classroom is the use of photomath. So I’ve seen schools that are putting the paint that turns it into a whiteboard on the top of a desk. And so the moment the kids come into math class, they’re working problems on the board, or they’re working problems on their desk and they’re going back to that old fashioned way of just working a lot of problems on the board so that we can see the process of how kids are working the math problems. Don’t let kids wait and do their math work at home. Do it in class, do it together. Let there be even peer process of working together. But try to get that process back into the classroom and shift for product as we’ve talked about before. Number three, let’s talk about what AI can’t do. Chat GBT is terrible on sources. Like, why would anybody say, give me five sources for my PhD paper. That is ridiculous. It can what they call hallucinate. I just call it fabricate, because hallucinate is kind of a term we use for humans. And so I don’t like to anthropomorphize AI. AI is an it. It is not a he, it is not a she. It’s very dangerous when we call it a he or a she because AI is just another tool to serve us. It can imitate us, but it is not a he or she, it’s an it. So we don’t humanize it, we don’t anthropomorphize it. We have to be really careful about that. But you can use a tool like perplexity AI, which is built on top of Chat GPT. You can use the Bing chat add in for Chat GPT. And we also have to remember that Google search is now generative. Bing chat is generative chat. GPT is actually built into Windows eleven, so you can chat with it in there. Bing chat is accessible in any web browser, and it does use the GPT tools. So Snapchat, which is used by over a third of middle school students in the United States, offered to write Write a poem for a student that I know of who complained about having to write a poem for homework, and it just offered to write it for that particular student. So AI is everywhere. You can’t just block it. And we need to realize that it also can’t tell on itself. AI imitates humans pretty well, and so the prompts keep getting better. The kids are sharing the prompts and how to prompt AI, and so we’ve got to be prompt, pardon the pun, that was intentional, and move into a new direction and have some pedagogical shifts on that. So let’s talk about what it can do. And I’m going to share ten uses quickly because I’m way over my ten minute time, but hopefully you guys don’t mind. So first thing that I have done very recently as I’ve tried to, I guess, grow up in a little bit in my use of AI is I’m making custom GPT. So I do have the plus version of Chat GPT. And so I have an 8th grade travel project that they make spreadsheets and they’re given a certain amount of money and they’re supposed to plan a trip and they have to have charts and they have to have formulas and they have to create a presentation that they’re going to present to their parents. And they’re also pretending that they are 18 and able to travel or whatever age they pick. Some of them pick 20 somethings. But I want suggestions for my students, but I want age appropriate suggestions. So I don’t want whatever tool they’re using to suggest, oh, go to this bar or go to any kind of adult type activity. I want it to suggest things that are age appropriate. So I created a travel tool programmed to be age appropriate where I programmed this particular project and programmed it to give advice to my students. It probably took me ten or 15 minutes, and I have blogged about it on my blog. Now, a lot of folks don’t know that you can use your voice to talk to Chat GPT. So I have the app on my phone. And so I brought up the custom GPT, the travel GPT on my phone, and I taught the students to talk to it so they could talk to it and say, I’m traveling to Sri Lanka. I like to be outdoors. I’d like five suggestions on what I could do in Sri Lanka. And then it will verbally answer them and they can talk back and forth. Now the nice thing about this is that I can see all of the chat that’s happened back and forth because the text version of the chat is saved on my phone, and so I can follow that whole thread. The one thing I do like to do is I like to ask the students to say, okay, this is Betsy, and I’m talking to the travel GPT. And so I know which student it is. I can usually tell because I know where they’re going to travel. So that was the first thing, and that was fantastic. I also programmed for it to do time zones, and here’s how that was really helpful. So while I’ve traveled a know, I could talk about Dubai, I can talk about Beijing, I can talk about Mumbai, talk about South Africa, but Sri Lanka, I have never had the pleasure of going to Sri Lanka. But the AI was actually able to give some great suggestions on things for the students to look for. And then the student was able to go in and do some searches on their own to kind of go in that direction. So in this case, we were using it as an advisor. And I’m going to tell you, these projects are so much better than I’ve ever seen before, because not only was I giving advice to my students, but my students were using AI. I did require the use of AI to advise them, and I required them to give me the citation for those chats or conversations. So the second thing is using voice. My students are programming in mad learn right now and making apps in 9th grade. And when they brainstormed about their apps and the content for their apps and the sitemaps for their apps, I required them to use AI as a feedback tool. Now, you may not know that Chat GPT can actually look at pictures, so I had them upload their logos into the Chat GPT that we were using that I had created for this, and it was able to give them feedback on those logos. And actually the feedback was pretty solid. It was some of the things that I had been telling students, and then they got feedback also on their graphic. So that was really cool and really helpful. Also helped them identify gaps. Now remember, we are not asking this tool to do any of the heavy lifting for the students. It is functioning as a feedback tool. And I will tell you that the presentations my students did on these apps were, I felt like they were college level. It just really advanced their capability of understanding apps because there was a content knowledge that the AI tool had on apps based on what had been programmed in it that was really solid and helpful for the students. The other thing is, because we had it talk, everybody in the group was able to listen and nobody was left out. And it was fantastic. And again, it was transparent for me and so I loved that. Number three, I’ve had my AP computer science principal students test AI tutoring. Now, our opinion, honestly is that it’s not quite there yet and a lot of the tools start asking for money pretty quickly on the tutoring. We do see a lot of promise, but it’s just not quite there yet. There is a blog post that they’ve kind of drafted collaboratively that I may work on and share, but AI tutoring cubot has been added, I think, to quizzes. There’s a lot of AI going in there and I’m really looking forward to the day and hopefully the spring or maybe next fall we’ll have some really good AI tutoring capabilities. They did test conmigo and there are some interesting things happening there, but again, I’m not sure it’s scalable quite yet, although it may be soon. And honestly, my email will probably be full of companies who tell me that it’s here and they want me to test their product and I’m happy to test it if I can. But that is going to be a great thing when we have that with tutoring, because here’s what we know is the students who score best on tests are typically making out tests for themselves and writing practice tests for themselves and taking it and self grading it. And very few students are going to sit down and do that for themselves. But this quizzing capability built into AI is going to really help. So I was testing with my son the capability of Chat GPT to voice tutor and we tried trig, okay? It was a big fail because the problem that I have with voice GPTS is I have to tell them to be succinct in their answers because they will just try to jump in and do the work for you. And I do not want AI to do the work for me until I tell it. I’m ready for it to do for me. So it’s overly verbose. Sometimes it is like that uncle who will never be quiet, who just talks the whole time and everybody’s like, seriously, you don’t understand. We’re just bored of tears. Okay? Sometimes it just will do that. And so you learn that you can touch the screen on the Chat GPT and tell it to stop talking and then talk to it a little bit more. But when I create my custom gpts, I tell them to give succinct answers and before it does creation to ask me if I am ready for it to create or invent. And I’m going to come back to that, because that’s a big secret for really getting some great stuff out of AI. But when he was asking it to teach him trig and quiz him, it would quiz, but then it would not wait. And there was nothing we could do to get it to pause 30 seconds or pause a minute and let him work the problem and then let him say his answer and check it. It just would not do that verbally. Obviously you can do that in text and do some of that quizzing, but we know that there are a little bit of limitations with Chat GPT, unless you’re using Wolfram alpha to help you with the math. And I wrote about that on edutopia about how to use math in AI and to preserve the productive struggle required in math. Okay, so let’s talk a little bit about art. My students are in 8th grade now, are using Photoshop 2024. Oh my goodness. It’s got generative fill, generative remove all of these generative features. And right when you open it up, there is a really cool seven step generative video on how to do all those things. Things. I do have about five computers that will not do the degenerative canvas fill. So you can take in a picture of like you at the beach, and then you can say, okay, I want to make this bigger on the left and the right, and then it will guess what it thinks Is there and give you some options. It’s really very cool. I’ll tell you. I was teaching this earlier this week, and I look up and I had like half the class with their mouth wide open. It reminded me of Michael in Mary Poppins when she said, close your mouth, you’re not a codfish. Because their mouths were just wide open and it was really funny. But that’s really cool. I love blowing their minds. And so that’s neat. There’s some really cool generative features in premiere pro. It’ll do audio cleanup like you wouldn’t believe. So we use that for editing movies, and it can take something you shoot on an iPhone and it can make the sound sound like it was recorded in a recording studio. So if you do film with kids, it’s really cool. There’s an AI tool called a script that with one button will let you remove crutch words or ums. Okay, Adobe premiere Pro can now do that for you. It can also remove gaps. So we made a really funny video and said, okay, anything that is more than a 0.5 2nd gap, remove it. And it took like ten minutes of film and turned it into a minute and a half. And it was hilarious because we were just playing with all these tools. So one thing I want to say is all the software you’ve ever used, keep updating it and keep experimenting with it, because as I’ve told my students, you can’t use it if you don’t know it’s there. So right now is a very important time to be updating your software, seeing what it does, and then experimenting with it, because that’s how you get experience with generative AI, and that’s also how you become the kind of person that a business or a classroom depends upon as we move into this generative age. Number five is there’s a really cool text to images, and while it’s in Photoshop and that sort of thing, canva has text to image. Bing now has doll e built in, and there’s a doll e custom GPT that you can use inside Chat GPT. And I have to mention that Microsoft paint has some incredible generative AI features. Yes, I said Microsoft paint. People are going to start talking about Microsoft paint because it comes with your Windows machine and it can do some of the coolest stuff, and it’s just right there. You don’t have to pay for it. So that’s one of those tools you want to update and you want to play with and use again. I have made myself a custom GPT I call it cat. So I say, hey, cat. And it’s a content AI task helper. And what I have programmed this thing to do is when I get in the car, I’ll fire it up. I’ve just given it a ton of stuff about the kind of stuff that I want to bring to you on my podcast and on my blog. And one of the big things is I want it to interview me to bring out the research based best practices that are already proven, but also the anecdotal experiences that those of you out there who are doing research and those of you out there who are innovating and iterating can try out. And I want it to be in a really practical way that anybody can understand. I have a lot of people who listen to this podcast who do not speak English as a first language. And so I want this podcast to be very accessible, and I want the blog to be very accessible. So I have it interviewing me. So I have about a 32 minutes drive on the way to school. So I’ve got some of these blogs that I’ve written on my blog. And so other day I was like, I want to talk about the importance of human relationships and AI, and also how I’m using AI in a classroom as it relates to how I teach and relationships. And then it asked me a question. It said, okay, what have you experienced in your classroom? And so this is one of those examples of where I’ve said, I want you to give me succinct answers. I want you to ask me short questions. I do not want you to create content until I tell you to. And it’s just like an interview. It keeps asking me questions. So I’ll have about 30, 35 minutes of interview of where I’m just talking. And it’s prompting me with actually some really great questions, some of it which I’ve given it already. And then when I get to school or when I get back home, I’ll say, okay, draft 800 to 1500 word blog post based on this content with this particular angle. Because usually as it’s interviewing me, I kind of start thinking differently about it. So it may draft it, then what I’ll do is I’ll actually look at it. And a lot of times the organizational structure is really solid. So I’ll usually pull it up on one screen, and then I’ll pull up my blog on another screen, and then I’ll kind of like, look back and forth and go, oh, I like this. I don’t like that. Then at the bottom of the article I am disclosing that I used a generative chat to help in the drafting of the post, and I will link to the interview that I did with AI, my AI chat tool, so that I can be forthcoming. And that way, if somebody really wants to dig a little deeper into what I’m talking about, they can actually hear me talk. Not verbally, but they can read what I said to the GPT. Now you’ll also see some little weird things that happen. So the other day I logged into and said, hey, Kat, and it keeps forgetting how I’ve told it to respond, and I don’t know why, but it does that. And then yesterday it called me Melissa. I do not know where it got that from. And I’m like, why’d you call me Melissa? My name is Vicki. So anyway, that was very strange. I have not published or shared this, but it makes me wonder if maybe somebody’s digging in the back end or played with it or whatever. I have no idea. I’m not putting anything confidential in there because it is helping me put my blog post out there. But I will say, for me, it usually takes me six to 8 hours to write a blog post. And since I only have a few hours every day, sometimes I’m looking at every once a week, sometimes once every two weeks. But it’s helping me generate content faster that I have written. So I’m using my human intelligence and it’s kind of like a digital assistant helping organize thoughts. In some ways, while I’m not offloading a lot of the cognitive work, I am, I guess, offloading some of the organizational because it takes all the answers and it kind of puts common things together, maybe like a database would. So it’s really, really helpful. And then I do link to that, and I link to it for many reasons. Some of it’s so you can learn how I’m using it, and others is just to kind of disclose. You will not catch me doing what Sports Illustrated did. Sports Illustrated, as most of you know by now, has been caught fabricating and making up people that weren’t really people when it had AI writing for things, and it was a complete lie. And ethics and integrity is so important. That’s a whole other topic. Okay, so the next thing, I also use AI to support programming. So I made a program. It’s a long name, like diva or something, and I let my students use that GPT. But I actually love Bing chat enterprise because I love to chat with pds. Now, you can actually do this in regular Bing, but in Bingchat Enterprise, you have that extra layer of confidentiality. So you could use confidential documents internal to your school if you have Bingchat enterprise to do this. But I basically can chat with a PDF. So I’ll say I’m teaching arguments and parameters and I need to pull out some of the standards for that. And it will go through the PDF. It’ll pull out those standards and give me page references for the pDF and then I can click on those page references and go to those pages. So if you have a massive PDF, you have got to learn how to do this. Now, I do have some stuff I hope to be sharing soon with some of my tips on how to do that. But for know, you can open up Microsoft Edge and load up the PDF and use bing chat to talk with it. You do need to give it. I have found just a little bit of time to quote, read it, load it all in the browser. I usually scroll all the way to the bottom, scroll all the way up, give it a minute and then start talking to. It’s very, very helpful. And people who use pdfs, oh my goodness, you are going to email me and thank me for that. Now there’s all these great add ins and there’s this really cool one for elementary teachers called coloring book hero. Again, you have to have GPT plus for this, but you might want to have somebody on your team who can do this. So we were going to play for the state football Ball playoffs for football. And so I had it generate some really cool coloring pages to celebrate that. I also generated some coloring pages for ancient empires and all kinds of things. And it’s really quick and it’s really easy and it’s awesome. So there’s all these tools that are out there, these custom gpts that you can subscribe to and use to generate not only text, but to generate really cool graphics. Okay. But my next one when I have to be out and have a substitute. This is one of my favorite uses for Chat GPT. I have a particular chat that I have trained on the format that I like to use for my sub lesson plans because I really like to give my sub a lot of detail. The time, the class is the name of the class, the roster for the class, step by step what they’ll be doing, a copy of, what’s in Google classroom, all that kind of stuff. And so what I’ve done is I have programmed just a particular chat in Chat GPT with that format. And I’ll go in and I’ll say I’m ready to do a subplan and what I’ll do is I’ll actually paste in the lesson plan I had written if I had been there and it will turn it into a subplan and I will just have to tweak it just a very little bit. That is such a helpful use. And then in my lesson planning tool we use chalk from power school. I’ll actually go to the very bottom and I’ll put like star, star star sub lesson plan for this page so that I have my original on there for when I roll those lesson plans forward. But at the bottom I’ll have a sub lesson plan that I generated that I can print out for my sub and that I had to be out with some eye surgery a while back and that was so useful. Now let’s shift a little bit to some, a little bit personal use, I guess. I use, I do use it to polish my transcripts. I do have Adobe Premiere pro now do my transcripts for me. It will pull it out of the audio and it is incredible. It does a great job. I have to do a few tweaks, but it’s really great. But I will pull it then into Chat GPT and I have programmed in my custom cat GPT that I can paste my transcript in and it will format it the way that I like to format it. Now let me remind you that you’ll want to use the copy box at the bottom so that you can copy a transcript or whatever. Last thing that I love is I have programmed my cat GPT also with a bunch of questions that I really like. I read one time that it’s the questions we ask ourselves that define our lives. And so I have a bunch of questions. I’ve been collecting questions now for 1520 years of a bunch of questions that I got from John Maxwell, old Tommy Zig Ziglar, and then my own questions that I like to ask myself. And so I can go to cat, my cat GPT that I’ve created, and I can say, and yes, I’m saying cat C-A-T my content AI task helper, and I can say, ask me a question and then it will pull one of those random questions to ask me. And not only is that helping me think about life and what I’m doing, and am I doing the things that matter or not, but it also very often prompts me in a direction for a blog post or something that I want to bring to all of you remarkable educators out there. So I love using it that way. So you can see that I’ve shifted from the parlor tricks and the jokes. And yes, sometimes I still use chat GBT to help me come up with present and gift ideas. It can be pretty good with that, as long as I’m not given private information. It’s pretty good with all that. But I’d like to conclude now with a few things. AI, like technology, is not neutral. It can be used for good or it can be used for bad. And AI and hi, human intelligence. It’s a multiplier effect. So if I use AI, I actually wrote some formulas on the board. I said, AI, if AI equals say, it’s a power of 100, right? It’s very powerful. AI, 100. But you have human intelligence of zero. 100 times zero is zero. AI will be worthless for you when you have zero content knowledge. But as you start moving up in your human intelligence, the AI can be even more useful for you. And that human intelligence, though, has to be more than content knowledge. It also has to be knowledge about how to use AI itself so that that content knowledge can be multiplied. So I am finding that I’m able to do more work, better work, more thoughtful work. I’m thinking deeper, I’m getting more accomplished because I am learning to effectively use AI where it’s appropriate, that I actually feel like I’m using more of my human intelligence than when I had to just take a lot longer to do things. I feel like I’m iterating faster. So human intelligence is vital to properly supervise AI. If you can’t properly supervise AI, AI can be a menace. AI can be a problem. We can’t just, people are concerned about rogue AI. Yes, there are things that need to be done to prevent that. And I am not your researcher person. I am a person in the classroom using these tools. The other concern I have is that we can’t let AI cheat kids out of an education. So the problem I’m not as much worried about schools who are having AI conversations and learning how to use AI in the classroom and are having writing department meetings on how to use AI. My concern is actually the schools that are trying to block it in completely, not even speaking about AI. Those are the schools, because they are still focusing on the product and it’s so impossible to catch without being biased against your weaker students and putting them in that box that they can’t get out of. So we have got to be talking about it. We’ve got to be how are we going to be making sure students understand the process of math? How are we going to make sure they know how to code? How are we going to revamp our writing in our classes? And the other thing is that there are methods that have always worked. Project based learning, oh my goodness. So helpful, so useful. And AI is a massive multiplier effect on project based learning. When you use it as a feedback mechanism, it’s really fantastic because you’re getting some content knowledge, you’re studying, you’re verifying. And I require students to cite not only the content, the original sources, so they have original sources of information. They’re also required to quote their AI feedback conversation and link to that. So I literally, at the bottom of each app page in the mad learn app that they’re building, have original links, hyperlinks, and then hyperlinks to the AI conversation so that I can go in and supervise the AI and see how they’re using it. This is really an exciting time. Peter Diamandis, in his book that I often quote, the future is faster than you think, says that in the next ten years we will have 100 years worth of change. And really, if you kind of start that ten years at November 30 of last year, I really feel like we’ve had five or six years worth of change. A lot of the stuff that I was talking about in January that I said would be happening in three or four years has already happened. And I think next year may feel like eight or nine years worth of change, and then we may start going faster and faster. That’s a lot of change. And so we all have to innovate like a turtle. As I often talk about it, more is taught than taught. And so we need to make sure that our kids understand how to learn about new things and that we are modeling how to understand new technology. We’ve got to be talking about deepfake. Deepfake has come to voice. Every family should have a code word that they use or that they know so that you can verify that that is another person on voice or FaceTime or email, that you can make sure that that is your family member so that you are not deceived. Because deep fake is really a problem where people can fake your voice or fake your. I mean, it takes 3 seconds now of your voice to fake you. So there’s so many things going on with AI. I appreciate you listening to this very long show And now I want to finish up by talking about our sponsor. So the Modern classrooms project will help bring you engaging, exciting teaching to your classroom with their free online course and community. I would love for you to go to coolcatteacher.com modern right now to sign up for the modern classroom’s essential course for free, where you will learn about these strategies, research, and resources that can drive student centered, self directed learning learning in your classroom. They can show you how to use technology to unlock deeper student relationships with your students. So many people ask me how to respond to artificial intelligence, and my response is always, as you just heard that we need to shift our pedagogical practice in the classroom to the time tested, research based best practices. That is what you get with modern classrooms. So go to our sponsor today, coolcatteacher.com modern, and sign signup. You’ll be glad you did, and thanks for listening. You’ve been listening to the ten minute Teacher podcast. If you want more content from Vicky Davis, you can find her on Facebook, x.com, TikTok Threads, Instagram, Blue sky, and YouTube at Coolcat Teacher. Thank you for listening.