Sean Tibor: Hello, and welcome to teaching Python. This is episode 107. It's all about artificial intelligence in the middle school. We're here with Dr. Nisha Talagala. My name is Sean Tyber. I'm a coder who teaches, and my. Kelly Schuster-Paredes: Name is Kelly Schuster Peritas, and I'm a teacher who codes. Sean Tibor: So welcome, Nisha. It's wonderful to have you on the show. I know that Kelly has been carrying around your book and her backpack for weeks now and telling me about it every chance she gets. So we're just thrilled to have you. Welcome to the show. Dr. Nisha Talagala: Thank you. Thank you very much. Thanks for having me. Sean Tibor: We've thought that this would be a wonderful topic to get into because as much as there have been a lot of recent headlines in the area of artificial intelligence and some breakthroughs that have happened in terms of accessibility of some really advanced artificial intelligence kind of broadly to the world. We think it's always an interesting topic for us to talk about how this affects learning and how it affects students and how we can bring more of this into the classroom. So welcome, and we're looking forward to a really great conversation with you. Dr. Nisha Talagala: Thank you. Same here. Sean Tibor: So before we get into the fun part of AI, let's start with the wins of the week, something good that's happened inside or outside of the classroom. And Nisha, we're going to have you go first because it's kind of fun to put our guests on the spot just a tiny bit before we get going. Dr. Nisha Talagala: No worries. I've been able to remember a little bit of last week, so let me see what I can go with that. So I would say maybe the win is I have a 14 year old daughter. She came up with a secret project that I'm not allowed to talk about. And since the Chat GPT API is not yet out, I've been looking for options. Right? Because that's what I would use if there was. And then I came across your article about Chat llama. Super excited. I'm pretty sure she can use Chat llama. Kelly Schuster-Paredes: It's all open source. That's awesome. Dr. Nisha Talagala: I'm in. So I think that is probably my win of the week, is that I've finally figured out how to help her with her project and keep it on the right side of crazy. So that's probably the win. The not the win of this week is I seem to have developed an addiction to fried chicken. I've eaten an extremely large amount of it during the week. I don't know why. And I guess you just stopped. You're making me hungry before my doctor gets involved. Sean Tibor: Yeah, your not so win might become my not so win very soon. Dr. Nisha Talagala: I think I found this place, and they have such good fried chicken. They just keep DoorDash from that place over and over and over again. Sean Tibor: There's a place by me called Fran's Original Fried Chicken, and it's one of those places where they've got the same neon sign in the window that's probably been there for 50 years, and it's probably still the same grease that they're frying it in. So it must be amazing. And I'm both excited and terrified to try it. Kelly Schuster-Paredes: My kids have figured out how to make fried chicken, so that's like, we'll cook tonight. And I'm like, oh, great. Fried chicken. I'm on the opposite side. I'm done with the fried chicken. Sean, do you go in person? Sean Tibor: Sure. I can share one. I had a bit of travel this week. I went to Cincinnati, Ohio for the weekend in February, which I know you're thinking, how did you pull off Cincinnati in February? What a great opportunity. Honestly, Cincinnati is a place that I lived for many years when I first started my career, so I'm pretty familiar with the city, although it's changed a lot. But the win was that my daughter was singing in a national honor choir that she got selected for there, and it was her first kind of big performance of this magnitude. And there were 215 elementary school students singing in the choir with my daughter actually 214 plus her. And it was one of those moments where I saw how hard she had worked to do that. She was right up front and center in the in the choir, and I was sitting right up front and center across from her to watch her sing. And when they opened their mouths and just this wave of sound came towards us in this beautiful concert hall that was built in the 1870s, I just had this huge moment of, dad, you cried. So proud. Yes, I cried. I admit it. It was a beautiful moment. The music was beautiful. These kids were just singing their hearts out. And it was really amazing to see such talented students from all over the country come together. So my win was very clear this week. It was nothing to do with technology and everything to do with just a beautiful moment, and I'm going to remember it for a long time to come. Kelly Schuster-Paredes: That's very sweet. I can see it. I knew you would be crying as soon as you started telling the story. He's the emotional one of the file, so I always want to share. Now about my course. I'm taking a data science course with Georgia Tech, six months boot camp, and it's kicking my butt, and I'm loving every second of it. And I want to say that my win is we started MongoDB and I was like, oh, wait, I'm going to pull up all my old videos from Michael Kennedy's Cohort about Fast API and MongoDB and actually learn it this time. Because when you did it the first time, it was a little bit over my head, but now we're starting to dive into it. But that's not really a win yet. So I think my big win, actually, this week in exploratory robotics is the kids are starting to pick up in chess again and they're like, we can play chess in robotics. I'm like, 100% everything to do with coding, problem solved, looking forward thinking, following patterns, looking for different alternatives to solutions. I was like, yes, we're going to do chess this week. And they were so happy and they were so involved. And we have a lot of chess players in exploratory, and I think when people walk in to visit the classroom, they're looking around, they're like, this is robotics. I'm like yes. Problem Solving 101. Go. So it was like a really good win, and the kids were excited that they had a change of pace. I still had my normal kids that are playing with a couple of robots and we're building automata kind of scenes with spinning dinosaurs with some of the younger kids, but most of them are playing chess this week, so it's kind of cool. Sean Tibor: Nice. Any fails this week? Kelly Schuster-Paredes: Tons. I'm having issues trying to build my histogram with a tuple in a list, but I'll get there, so it'll happen. Sean Tibor: We just got to crack the code. Kelly Schuster-Paredes: Just trying to dig into this data and I'm just like, Damn tuples. Sorry, I will get there. Sean Tibor: Well, my fail this week was a fail up until about an hour ago, and it was just some code I was trying to work through and I kept getting this weird error and it was all in JavaScript, which I'm not super familiar or comfortable with, so it took me a while to get there. And once I figured it out, it all made sense and I was able to work around it. But I had a couple of hours where it just was not working, and I was getting cryptic error messages, and I had that moment of realizing, oh, yeah, this is what it feels like to be back in the beginning of learning a language again and not having that immediate knowledge of what's happening or how to fix it. So it worked out really well in the end and it's running beautifully, but it was a nice little humbling moment that I think I kind of appreciated when it happened. Kelly Schuster-Paredes: I love that because it's a good reminder for us all. That learning process. Dr. Nisha Talagala: Bugging is joyful. Something, being able to get it to work and then having it work. Sean Tibor: Yeah, and it's like, it's that breakthrough moment where you just have that rush of like, yes, it worked. And I was so far behind. I think I Googled at one point, where do I find the line numbers in a JavaScript stack trace? I couldn't figure out exactly where to go. And then I got it working and that helped it todd, but it was a bit of a tough sledding before I got there. Kelly Schuster-Paredes: I'm glad for that. And then you got there like always in the end. It's awesome. Persistence is I want to dig into this book. Sorry, I'm like, let's go do it, because in two weeks, I'm starting an AI unit. So I'm going to let you go. And Sean, you could start with no. Sean Tibor: I mean, I just love this. This is so Kelly. It's like I have this unit coming up. So two weeks ahead of time, I'm going to get a guest in here. We're going to talk about AI. I'm going to have a head start on it. Kelly Schuster-Paredes: She's going to write my lesson plan for me. I'm done. Sean Tibor: She is so much better prepared than I am with all of these things. So this is the stuff that I enjoy the most about working with Kelly is this kind of leaning forward approach that she always has. So, Nisha, we wanted to just start by introducing you a little bit. You've got quite the background and training and knowledge to bring to bear on this subject. And I think it's probably one of those things, I'm guessing a little bit here, that when you first got into your career field, there was really no way of projecting this far ahead to see where things like machine learning and AI would go. But you ended up being kind of uniquely positioned for this in the beginning. So I would say just to summarize a bit. So you have your PhD in computer science from UC Berkeley, which is a phenomenal school, and I still always think about the book by Cliff what's his name about Inside the Cuckoo's Nest, where he was chasing down KGB hackers from the UC Berkeley Astrophysics Club. It's an amazing book, but I've always had a soft place in my heart for the UC Berkeley computer science programs. As a result, it's a phenomenal, world class institution. But since then, you've also created an organization called AI Club. You've been the founder there. You've also started AI startups for banks and large corporations to help them build AIS. And that was successfully acquired. You've also organized conferences in AI. But I would say probably the thing that is the most relevant here and the thing I was most excited about is the work that you've been doing with your own daughter to help her learn about AI and what's possible because there aren't a lot of materials out there for students of her age to learn about AI. Absolutely. So how did I do? Was that a pretty decent professional summary? Dr. Nisha Talagala: Awesome. Thank you. Yes. Sean Tibor: Anything I missed or anything you'd like to add to that? Dr. Nisha Talagala: No, I think I'm good. Sean Tibor: Okay. All right. So, Kelly, I'm going to let you have the first question, and you're on mute. Kelly Schuster-Paredes: I got a dog in the background, and he loves this time of night, so it's always good for our live streaming. So we'll skip right ahead because you've done so many things in AI. And I just want to dive right into the book because I think our listeners will be interested in a lot of that and I know you wrote this book. You and your co author wrote this book kind of with the mindset for your kids. What was kind of like your thoughts on that? Start us off, like, go back. You made this amazing book. Dr. Nisha Talagala: I think the book sort of grew out of our teaching experiences. So I started by trying to teach my daughter AI when she was nine. That was five years ago. And I created a first project for her. And I realized a few things. One is actually it's very engaging, but then she ran out of things to do very quickly because she ran out of data. And I found that getting data for her was boring for me and boring for her. And then the thing kind of stalled, right. But I was kept it in the back of my mind and stuff like that. And then we started teaching her a little bit. Then we started teaching other kids her age. And over time, we developed sort of a set of curriculums. Yeah. And things like that. And we kind of started maybe in sort of the way you're we have a class, it's coming up in two weeks. What are we going to teach? Came up with something. Oh, that worked. Oh, that part didn't work. Kids were bored, you know, stuff like that. And then we got it better. And then we had a pile of curriculums that we started, you know, sharing with other teachers. And then we sort of realized that maybe putting a bunch of this stuff into book form would be helpful. I personally love books. Even though I do get a lot of stuff online, I still love books. I like having it in book form. And the thing with the problem with AI books out there is that they're either really simple or really kelly. Complicated. Kelly Schuster-Paredes: 100%. Dr. Nisha Talagala: They start off really nice and fun, and you're like, okay, I'm in good shape here. And then suddenly the word Bayesian Statistics comes up and you fall into off some cliff, and you're like, I have no idea what just happened. And some of you never seem to be out of that cliff. Kelly Schuster-Paredes: Go ahead. Dr. Nisha Talagala: No, please. Kelly Schuster-Paredes: No, I was going to say I'm going to stop right there, because if you started the book, you would think, like, you open this book and right away I said this in our episode, our previous episode when I was reviewing the books. You start off with this book and what I love about this, and I don't know if you did some AI studies or if your publisher helped you out, but the way that you chunk the reading, I've told all my students, I was like, go read this book. It says it's for teachers, but you're smarter than them anyways, so just go read the book. And it's in like, these little paragraph chunks, and you really just go in straight away. Like in page 16 about anomaly detection and object recognition and speech recognition. And that's great. So continue. Sorry, I have to add on that. I have to visualize this idea. Dr. Nisha Talagala: Well, I guess one thing is that we basically just opened up a bunch of textbooks, right? And we're like, what do these books look like? The first thing we noticed is there's a lot of pictures for every unit of text. So we said, okay, we need to have a 50% text to picture ratio. And we weren't there. And then we went back and just redid everything and said, okay, we need to add some pictures here and pictures there. We had an illustrator draw stuff for us and stuff like that. But yeah, so that was sort of one of our goals, was to just sort of you can go a really long way before you fall off a cliff, and you apparently never have cliff unless you really want to. The problem is that it depends on the order in which it's presented to you. Sean Tibor: Well, and I think there's a lot of concepts within AI and especially within machine learning that are really hard to grasp and really hard to kind of access with your cognitive skills without pictures. Right. So I remember we were doing a workshop, Kelly and I, a few years ago with machine learning, and I realized that the professor had explained that whole idea of a gradient curve in such a beautiful way with illustrations and with some annotations on it to kind of point out different parts of it. And what I realized 25 years too late is that he was describing a really practical use of calculus that would have been really helpful to me back when I was actually taking the classes, because it became alive, right. The visualizations made it live, whereas the equations by themselves or the dry explanation couldn't do that. The pictures made it come to life. Dr. Nisha Talagala: Yeah. This is really I don't know. So when I was young, nobody bothered to tell me why math mattered. For some reason, I liked it, so I kind of got away with it. But kids these days don't put up with that kind of stuff. They will come and ask you, and I'm doing this why? I could be doing 300 other things instead of doing this. Can you explain to me why this matters? Sean Tibor: And even just that accessibility of knowledge, too. Right. You're in a marketplace of teaching now. You're not the sole source. You don't have a monopoly on it as a teacher. So you can either lean into that and say, that's great, here's some other places for you to learn from, or you can try to resist that to your peril. I think that's a really good way of kind of highlighting that is when they come to you saying, why do I need to learn this? You're competing. Kelly Schuster-Paredes: Yeah. Dr. Nisha Talagala: And it's actually one of the coolest things about AI is that it is really easy to connect AI to the map. Kelly Schuster-Paredes: I was just going to say that as you dive deep and I'm like and again, mind you, I am now just a coder of five years and a biologist in the which is on the way on the obviously on the way on the other side of actually using technology at my time. Now they do it all the time. But as I'm digging in here and looking at the math, I actually used four pages of your book today because we were doing yeah, I'm telling you, it's like, buy the book, people educators, buy the book. It's amazing. We're doing a unit on big data and again finding the data for 8th grader to get, and we've learned requests, and I could go into Scary Sean and they could do all this stuff, but we don't really have that much time, so I just want them to get the data and tell the story. And you go into the storytelling, and we're not even talking going into the linear regression and all the other types, but just going into the fact of what is your story and everything. I don't even know what the question is. I just love your book. Hold on. Sorry. But when you're going through it, you were thinking about this. I guess you got the fundamentals of where you came from and how you solved problems. Is that similar to think? Dr. Nisha Talagala: Yeah, I think maybe one thing that benefited myself and Cindy is that we have used AI in the real world. We've built it, we've used it to solve problems, and we also know how people use it. So being able to sort of simplify that and provide it to people, AI is not really what matters. It's the fact that it's useful to solve a problem is what matters. And half the battle is, okay, I'm excited about AI, but maybe the problem isn't right for it. Kelly Schuster-Paredes: Absolutely. Dr. Nisha Talagala: Do I even know what my problem is? Because AI can't help you if you have no idea what your problem is. Sean Tibor: How did you include Misuses of AI or misapplications of AI? Because the real world is an incredible teacher for that. Right. Showing you opportunities where people have used AI when it probably wasn't appropriate or wasn't ethical or whatever it could be. How did you incorporate that into your thought process? Around the book? Kelly Schuster-Paredes: Chapter two. Dr. Nisha Talagala: Thank you. I actually don't remember. Kelly Schuster-Paredes: I'm kidding. I was just making that up. Sorry. I liked it. Dr. Nisha Talagala: What I find is that sometimes in a real world, examples and particular examples that kids can relate to can go a long way towards that. There are some examples that are hard for kids to relate to just because there are problems that adults understand. But I can kind of, like, start with sort of a simple thing, which is the interaction between AI and your privacy. So you can ask kids a bunch of questions like, okay, if your grocery store or your shopping website knows that you like eggs, is this a problem? And most of the class will say, no, it's not a problem. There will be a kid who says, my interest in eggs is none of their business, stuff like that. And then you ask, okay, what if they know that your dad gets medication for back pain now? They're little. I'm not so comfortable anymore. What if they decide to tell your dad's boss that he has back pain now? Pretty much the entire class will say no. And so sometimes just by letting them think through it in a way that they can relate to it, they can kind of understand that there's no black and white and there's a lot of subtlety. And what really matters there is their opinion. There's no hard and fast place between. Most people will be okay with almost everybody will be unhappy with the back pain disclosure. But somewhere in between is your comfort level. Kelly Schuster-Paredes: Absolutely. Dr. Nisha Talagala: And it's probably important to know what it is. So stuff like that. I have them do exercises. I ask them questions. It's amazing how much bias there is out there. If you just look and then all. Kelly Schuster-Paredes: Seriously seriousness, you touch on topics that I think people who are not in technology don't really understand. They read the headlines and they go chat GPT and fake news and deep fakes, and they don't really understand the words that they're necessarily saying or how to teach it. And that concept of going through these little problems and the unplugged exercises is a really neat guide for teachers. Did you have the mindset to hope that this gets into more classrooms, more schools? Was that the mindset of your book? Dr. Nisha Talagala: That's definitely one of the reasons why we decided to put it in a book is because we felt the reach would be much greater. And by working with students and by working with schools, we kind of realized that you need to have sort of every type of exercise. Unplugged activities are just so helpful because teachers don't want the kids to be on computers all the time. You really want them to be able to shut down their computer and actually think and converse and discuss. And so it's important for that reason. And then when you go to Asia and things like that, there's a lot of lack of WiFi. There's no stable electricity. So sometimes the unplugged is all they can do. So we try to create enough that it can be accessible to a lot of audiences. But at the same time, AI being AI, there are things you can only do online. You want to have a lot of online activities, but you also want to have every other kind as well. So, yeah, our goal is absolutely to get it as far out as many students and teachers as possible. Kelly Schuster-Paredes: Awesome. And then I guess the other question what are your hopes for keeping this book current? Because with Chat Llama coming out already after Chat Gbt, and then the next thing and the next thing and the next thing with it being open source, what are your kind of hopes that how long is the longevity this book, and when are you going to write the next one? Kind of thing. Dr. Nisha Talagala: It's a really good question. And this is something that Sindhu and I talked about and talk about often. So I think there are kind of two dimensions to this. So this is volume one, independent of your question. We are working on a volume two. Volume two is a more deeper because it has a lot of stuff, but it doesn't have all the things that you might want to cover. The other fact of how do you keep the books current? The best thing that we've thought of is that we are thinking of including, including online supplements to the book fairly often. So if you get the book and you can just go and say, okay, what are the online supplements? Because the book definitely will get old relative to tech, and it's not practical to necessarily ask people to keep buying next month. You have to buy a new physical book. That's not practical. So I think we'll just basically release the book every once in a while. But we'll have online supplements excellent. That we can add on to all the time that people can then just go and look, hey, what's new? What else should I know besides what's in the book? Kelly Schuster-Paredes: So I'm going to totally monopolize all the questions. And so for educators out there, for me, I like to pick a book. I'm like scan through it. I'm like, oh, this is a great topic. I'm going to pull some resources from that. Some people like to sort of go from start to finish. How long does it take for maybe I know you probably don't cover maybe exactly like the book because you have your own curriculum, I'm sure, in your club. But if someone was supposed to do this kind of curriculum book as a guide, how long do you think it would take? Or how long did it take your daughter to get through that much? Dr. Nisha Talagala: So one thing is that we actually publish four curriculums just for the book. So if you want to get the curriculums, they're actually online and they match the book exactly. And I think there are QR codes in the book for that. If you're teaching this entire book top to bottom, I think it'll take about a year of instruction, two terms basically, to cover the book. Kelly Schuster-Paredes: So that's a hint out that you need longer than I'm going to get in trouble for saying this. You need longer than nine weeks to teach. Python. Dr. Nisha Talagala: Yeah. And then the other thing I think is, depending on what you're trying to do, if your goal is to try to give them an introduction to the technology, help them appreciate the ethics. Maybe they don't code as much, then that's one way. If you want them to really get into the coding, which I strongly recommend, then that, of course, will take longer because they really have to wrap their brains around it. But yeah. And the book is about a year's worth of nice curriculum in there. Kelly Schuster-Paredes: Do you want to jump in? Sean Tibor: Sean, I wanted to switch gears a little bit because I was reading your Forbes article about the three E's, and I really like the way that that was put. And it's something that I've, in my professional life, have been thinking more about as well. For example, this week we're writing our annual objectives for my team. So what are we putting out there as our goals for the year, the things we want to accomplish? And I have two relatively junior engineers working on my team now, and one of them is fresh out of her master's program and just started a few weeks ago. And so we're writing our objectives. And I could see that she was kind of stressing about how to formulate the writing and everything. And I said, okay, well, just get it kind of close enough, and then put it in Chat GPT and tell it to rewrite it. Right. Dr. Nisha Talagala: Fair enough. Yes. Sean Tibor: And she had that moment where she was like, Wait a minute, I can do that. That's okay. And I was like, well, you have to balance your effectiveness and your efficiency as an engineer with your time. Right. There's nothing that says that you can't use this for it. We're not putting anything in there that's particularly sensitive information or something that's a corporate secret. So why not use the tool to help get yourself a little bit further ahead or give yourself a new perspective on how to write these? And so when I read your article to prepare for this, it really struck me that those three ease of the ethics of it, the effectiveness and the efficiency of it, really struck me as a great framework for evaluating not just Chat Gbt, but other AI tools and other tools that are available to engineers. And I just wanted to kind of get a little bit of a summary from you on how you thought about that, how you write about it, advice that you would give to new people coming from that student mindset into the professional world, how to think about using these tools to increase their effectiveness and how to use it to learn better. I guess, as they're making that transition, sure, yeah. Dr. Nisha Talagala: Both in AI club as well as in my past companies, I also have employees at various levels that I've mentored. And one of the things that I think is a big transition for particularly a fresh grad, is unless you were doing a thesis of some kind, like a master's thesis or a PhD thesis, unless. You did that. Your entire college life is essentially doing things people told you to do, usually very precise, very structured things. And if you did them well and on time, you got good grades, and you're a I feel good. Right? When you go into a company, if you're lucky, your first few jobs will be like that. Someone will tell you what to do. You don't get to say, and whether you do it or not, you do it. You feel good, everyone's happy. But as you become more and more senior, it really comes down to, are you solving a problem? And sometimes that problem isn't even technical. One thing someone taught me when I was very early in my career, I think my first boss taught me, is that ultimately you are paid to solve problems, and the more senior you are, the larger vaguer and weirder your problem is. And fundamentally, you're getting paid so that the person paying you doesn't have to worry about the problem. The only reason, if they end up worrying about the problem, that's pretty much your first clue that you're not doing your job well. So it's a leap for them to realize that sometimes that means they have to figure out the problem. Kelly Schuster-Paredes: Sometimes. Dr. Nisha Talagala: Sometimes it isn't even well defined, right? They have to figure out what part of it is hard, and sometimes it's not the technology. Sometimes there's a guy over there that won't agree with me, and my real job is to convince him, and that's effectively what I need to do. And so that's that leap and feeling the freedom to make that decision, realize what the problem is. Realize that you're not being mandated to do it a certain way. Part of your job is to figure out how to do it. It's a leap, and it can cause a lot of stress for young engineers, and they encounter that step somewhere in their career, and it causes a lot of stress. And you just kind of have to give them examples of how other people have done it or examples of how they could do it to help them realize this is expected, this is normal. It's even okay to fail in doing this. But it is critical that you reach some comfort with the vagueness. And the reason I put the three E is because ultimately, those are the only three things that ultimately matter. How you do it doesn't matter as long as it's effective, it's efficient, and it's ethical to break the law, you'll be in trouble no matter what. Kelly Schuster-Paredes: I love that. I love that. I actually was having a conversation with an English teacher today because we were talking about Chat GPT. And for English teachers, chat GPT is something that's hard hitting foreign language teachers. They've been dealing with this problem for a long time. Google Translate has corrupted their kids because one way or the other, and Calculators and photo math has disrupted math teachers but for English teachers, Chat GPT speech not speech, but word recognition and natural language processing is like it's a new thing. Even though they've had word correct for many years and we had this conversation and it kind of hit me and maybe you can tell me your thoughts on this when I said effective, efficient and ethical. And the teacher said to me, yes, but that's for a technical person like yourself. What are your thoughts on that? And I'm like, but you have to be efficient, effective and ethical. Why waste our time writing a five paragraph essay when Chat can fix it for you and you can make it better? Any thoughts on technical? People are ethical, efficient and effective and English teachers are not. I mean I'm just kidding. He's not that. Dr. Nisha Talagala: Maybe I can kind of maybe talk about the three E is in the context of non technical and then maybe talk about the specific challenges that English teachers are. So it absolutely applies outside of technical. And one of the places, I think, where engineers have a hard time is when they get more senior, their job ceases to be completely technical and that's where they usually get into the kind of trouble that they're not used to that school didn't prepare them for. So it applies everywhere, and it possibly applies even more in the non technical because being effective in a nontechnical role requires an amazing array of soft skills that are really hard to pin down. But when someone's got it, they've got it. I think it applies. Now, coming back to the English teacher, I think part of the problem is, and I really have a lot of sympathy for these guys because the problem is sometimes they are not necessarily allowed to not teach things. Like if an English teacher were to say, okay, you know what? I have decided that my students don't need to write essay, learn how to write essays. Right. I would like them to critic Chat CPT instead. I don't think they have the authority. The curriculum, the state standards do not give them the authority. So they're in this kind of box where the standards determine that they are expected to teach X. And now it's difficult to teach because the kids are all using Chat Tpt to do it. So I think they're just kind of like stuck there between I'm curious what you guys think, but I think that just makes it just so much harder to maneuver. Kelly Schuster-Paredes: It's a difficult conversation because it's twofold, right? Most of the arguments on the I wouldn't say anti Chat GPT because it's there, so there's no not happening kind of world. But the two arguments are the ethical sidepoint of a student, et cetera, et cetera. But the other side is this worry that we are developing people who can't think for themselves. It's a weird predicament for a lot of people, and especially as an educator in technology who's pro these problems. And the idea that an AI just causes more ability to solve problems. You just have to find out what are our new problems that we are searching to solve? And I think that's a mindset that's difficult. Dr. Nisha Talagala: Yes. I don't know if you guys saw, but I wrote a Forbes article on. Kelly Schuster-Paredes: This. Dr. Nisha Talagala: Chat ept and imagining human intelligence. So I think to your point, I completely agree with everything you're saying. I personally don't get too hung up on is it ethical to be using it? If there's a tool and the tool makes your life easier, the problem becomes when you keep the problem the same. And now the tool solves the problem. If you make the problem different, then suddenly your brain is still engaged. It's engaged doing something bigger than what it could do before. And so now, if you look at like one of the things I've learned by teaching AI to kids is every year we run a research symposium where advanced middle schoolers and high school kids present research activities. Last year we had somebody who did detect forecast charge for EV batteries, someone who predicted whether you were doing a tennis serve correctly, some lady who did something involving the thyroid cancer, a bunch of stuff. And these are things the only reason kids can even approach these problems and apply their imagination is because of the level of tech. If you insisted that they write every piece of code by hand, none of these kids could ever have done this. And so it's really a question of being able to give them bigger problems. But that also means that the teachers have to be ready to imagine bigger problems yeah, it's bigger problems and evaluate bigger problems. Sorry, you were saying? Yes. Sean Tibor: No, I was saying it's prompting so many thoughts. And I think it's getting back to that not even the problem solving. When we talk about humanities, we talk about social sciences. I mean, there's certainly problem solving in there. But I think it really comes down to the what are we really trying to teach? What's the outcome that we're trying to engender by taking this approach? Why are we writing the essay? Why are we writing the poetry? Why are we making the drawing? Why are we doing the sculpture? Right? It's not necessarily about the final product. It's about the process that we go through to produce that right. The creativity, the act of creation. And so as I'm thinking about what these tools can do and what they can provide for us, what they provide are maybe ways to avoid the parts that we're not good at, so we can focus on the things that we are good at, the things that we are creative about. And I think that the part that's challenging for most in these spaces is up to this point, they've assessed that creative process by the product that is created by the finished product, not by the and with some formative parts, some parts that do assess the steps along the way. But I can also empathize, if I'm an English teacher, reading, you know, 150 kids poetry submissions, right? How much am I really taking the time to understand their voice, their their point, their the style that they bring to it, their their own unique contributions to it, versus kind of scanning through this one, making sure that it fits the basic criteria that I have in my head or that I've got on a rubric and giving it a grade and moving on. So there's two sides to this. Like, to your point, if what we're trying to create here or what we're trying to encourage is the creative process and this act of making things, then we need to continue to find better ways to assess and give feedback to that creative process. And that means we're going to have to shake up a lot of the ways that we've been assessing things for a long time. And Kelly's, I know given this a lot of thought around assessments and how we give students feedback so they can learn better, but that's the things that are striking me right now is that sure, some kid who's got a deadline tomorrow to produce an essay, five paragraphs on something that they really don't care about and they're not interested in. Of course, Chat GPT is a great shortcut and they're probably going to take it. But if the assignment or the creative process is to really learn about something that they're excited about and interested in and find a way to write something new and different about that, then why would they choose Chat GPT when they can write it themselves and make something new? Dr. Nisha Talagala: I completely agree with you. I think basically the model that it encourages is a good one. A model of being creative, being imaginative, being unique. It just unfortunately, flies in the face of almost every kind of assessment that is done in the world. And some of those assessments, the reason that assessments stay the same is because you want to compare ten years of assessments that's really, really hard to do in customized, project based learning. On the other hand, if you look at, like, if you go back to the the threes companies don't hire you based on your test score, after a while, nobody cares what your test score was. They want to know if you can solve a problem. You probably remember coding interviews where someone would ask you about the syntax of code. At least these days, I don't ask anybody about the syntax of code. If you cannot use Google to figure out the syntax of your code, I don't want to hire you regardless of what you remember. Right. Sean Tibor: I want to hear what problem you're solving. Kelly Schuster-Paredes: I'm turning off lead code. No more lead code. I'm done. Sean Tibor: But that's really what it comes down to. If we're looking at what is the value that people bring, not just the employee, but the sculptor, the artist, the creator, the writer, the coder, the engineer. What's? The value that they bring is their unique ability to solve problems in effective and interesting and efficient ways. Right. If that problem is, I have to get this creative thing out of me and I don't know how to do it, then why not use the tools that you have? Why not bring that forward and use the tools that are out there? Whether it's chat ept or generative art or something that's new and unique, it's another tool to help bring out that creative part of our human spirit. Dr. Nisha Talagala: Absolutely. Yeah. And then as long as you do it in an engaging way as opposed to a push button way, it will bring out a side of yourself that may not be possible without the tools sort of making things go faster. Kelly Schuster-Paredes: And I do have to say that the idea of creating the questions to your own problems, especially in the educational world, is huge. I mean, today, for example, I gave them the data that they collected on themselves, social media, distance, whatever. And the original idea was, go find some CSVS and tell a story and make some plots and practice Matt plot live. But the idea then turned in after I put the book up, we read it out loud in class and we read chunks of it, and I was like, we're going to do storytelling. I keep talking about storytellent telling in my class, in my own personal boot camp. What are your questions? Come up with eight questions. And I'm not sure that the kids have many opportunities to come up with their own direction. And the questions were phenomenal. Kids that aren't great coders aren't like the shiniest apples with the python code. All of a sudden, we're like the magicians. They were wielding stuff, and they're like, oh, and what about this and this and it's that opportunity to just let them think. And it was brilliant. Dr. Nisha Talagala: One of the things I love about teaching this age group is that somewhere around between eleven and 18, there's this period where their brains have really kicked into high gear, and the fear they develop as adults has not yet kicked. Kelly Schuster-Paredes: In 100% during that time. Dr. Nisha Talagala: They are fears, they're imaginative, and they are completely, you know, without any kind of like, oh, my God, I won't be considered cool if I ask. No, say something. And they come up with the most amazing ideas. And I love the storytelling because one thing I've noticed kind of working with kids, is that it's really helpful for them to understand how to speak about what they know. And one of the reasons why we run this research symposium is to help kids learn how to speak. And one of the things that happens very often is I will be mentoring a kid, and he has ten minutes to describe. He will spend seven of them describing something he found really annoying. And I was like, okay, I get it. I appreciate it. I'm willing to listen to you, but just the fact that you found it annoying doesn't mean it's interesting to someone else. What do you think your audience cares about? Not this, but it took me so much time. I get it. I will listen to you focus on stuff that they're just helping them to sort of understand what the world around them needs to know is a huge lesson. It'll just help them become better communicators over time. And it sounds like your students have figured out the right questions, because half the battle is the right question. Kelly Schuster-Paredes: Yeah, I always say that. Sean Tibor: I was going to say one of the things that I think Kelly showed me how to do that worked really well and was very effective was if they have trouble finding the right questions, encourage them to ask the wrong questions first. Right. Get those out and get the bad ideas out of the way. And it's amazing how effective that is at getting those creative ideas going, because they might ask a question that they think is dumb or irrelevant or not a good question to ask because they've already been trained what good questions are and get them to think for a moment. Wait a minute. That's actually a great question. Right? Like, you didn't think it was something worthwhile, but it's a really important question to ask. Let's see how we could solve that. And suddenly their whole view changes, and they go from being kind of skeptical or apprehensive about getting out there, and now they're engaged. Now they're bought in because something that they came up with is valuable and real. Dr. Nisha Talagala: Yeah. There's a tremendous opportunity for everyone to sort of find their passion and find their unique idea and see how it applies. And it's very engaging when that happens. Kelly Schuster-Paredes: 100%. Dr. Nisha Talagala: Yeah. Absolutely. Kelly Schuster-Paredes: Well, all I have to say is, when I first started coding, I told Sean that Matt Plot Lib was my favorite library. And I remember, like, five years. I was like, I love Matt plotlib. It's so easy. It's so wonderful. Then I got into data science, and then they told me in the course curriculum that machine learning was coming next. And then I found your book, and I wanted to share it with my colleagues not my colleagues, my cohort, who are all adults. It's phenomenal. It just goes through and it puts all these pieces together of for me, a lot of people ask, Why Python? Why do you teach middle schoolers, Python? And I tell the students all the time, and Sean told it. It's literally and now we have Chat GP to say yes, and it's Chat GPT, but it's literally the world around us, and it's a powerful language and an amazing book to go along with the powerful language. And I showed a parent that came in today was like and look, you can do it all with just two pages of code. Somebody was really excited about that. Dr. Nisha Talagala: Thank you so much. Really glad you're in. Kelly Schuster-Paredes: Thank you. Sean Tibor: I'm going to steal it from Kelly if I can ever get it away. Kelly Schuster-Paredes: Just by your own bulk. Sean Tibor: Okay. Kelly Schuster-Paredes: I've already written in this. Sean Tibor: I'm never getting it at mine. Fair enough. Well, I know we're running towards the end of our time together, Nisha. Is there any other resources that people should go check out or places where they can keep an eye on the work that you're producing with your partners? And coming out of AI Club, what's the best way for people to keep abreast of everything that's happening? Dr. Nisha Talagala: So I think there's probably two places. One of them is that AI club has a resources page. There's a ton of stuff in there. We release exercises, code. We have about 400 data sets for kids to go tinker with. Kids safe data sets. Kelly Schuster-Paredes: I have just a mental note going on the weekly overview tonight. Thank you very much. Sean Tibor: You had me with Kids Safe data sets. I like that. Dr. Nisha Talagala: Yeah. And then we write blogs every once in a while. We just launched a Chat GPT course for teachers, and there's a five minute video on our website where I talk about it. What is this thing? Where did it come from? Why should you how should you think about it? And I also have a cute poem that Chat Chippy wrote for me explaining chemistry to an 8th grader, which I thought was just very sweet and kind of represents to me all the positive things about that tech. So, yeah, so the resources page and the blogs on the website are a great place to do it. I write in Forbes about once a month, sometimes a little more. And I write about not just education and AI, but just all topics AI. So that's another place to kind of look. Those are usually the places like this stuff. Kelly Schuster-Paredes: And you have a wealth of projects, of real world projects. I pulled up and I think Google Scholar and all kinds of amazing things you did before and very interesting things that takes people to read about after they've read your book and learn a little bit more about AI. Dr. Nisha Talagala: Thank you. Sean Tibor: Well, I want to say thank you for joining us tonight. It's been just really enjoyable to chat with you about all of this, and it's an area that we're both excited about and interested in. It's just lovely to hear your perspective and the thoughts that you're sharing and the resources that you've made available to everyone. So thank you again for coming. It's been a real pleasure. Dr. Nisha Talagala: Thank you so much. I really appreciate it. I really enjoy the conversation. Thanks for having me. Sean Tibor: Kelly, any announcements for our audience this week? Kelly Schuster-Paredes: I'm supposed to plug I forgot. Innovation Institute at Pinecrest. We're hosting Innovation Institute and I'm trying to remember the dates. It's in April. My boss is going to kill me. It's after April 13. It is. I'm going to tell you, tell you, tell you, tell you april 17 and 18th, and it's on our website at Pine Crest. And then also python you want to talk about that? Ed Summit? Sean Tibor: Yes. We have the Education Summit happening on April 20 this year, so that's the Thursday when there are tutorials going on on the premises there. We've got the same room as last year. There's plenty of space for everyone, so please sign up if you're already registered for Pycon. There's no additional fee for registering for the Education Summit. We will probably have a different take on kind of remote attendees this year. We weren't able to get live streaming from the room to our remote attendees, but I know we're working on some ideas for how we can engage the community on that day and make it a big python education day. Whether you're there in Salt Lake City or whether you're dialing in remotely a way to just make it a celebration of computer science and python and learning. So looking forward to that. If you want to get more details about how to register for that, I'll put the link in the show notes for it. If you have questions for Kelly or I or for Nisha, you can always submit that through our website at Teaching Python FM. We're also on Twitter at Teaching Python. I'm at Smtyber and Kelly is at Kellypared on Twitter. I don't have any new social media networks that I've joined in the last few weeks, so no new handles to share. But if you want to get in touch with us, we're always excited to hear from our audience. Please feel free to reach out to us. So I think that does it. So for teaching. Python, this is Sean. Kelly Schuster-Paredes: This is Kelly signing off.