Track 1: [0:21] Hello, and welcome to the Teaching Python podcast. My name is Kelly Schuster-Paredes, Track 1: [0:25] and I'm a teacher who codes. And I'm without a co-host, but that's okay, because I'm here with Mahmoud Harding. We're here to talk about how, oh my goodness, everything. We're going to hear you talk about Python and data science and math, oh my, you know, whatever else. We're going to go down like the yellow brick road of fun here, I think, and nerd out a little bit. Thank you for coming. Track 2: [0:50] Thank you for inviting me. I am a longtime listener and a first time guest. I feel like I'm fanboying over the Teaching Python podcast. I've listened so long from the first time when I met you a few years ago. And I really enjoy the conversations you have, especially coming from the perspective of being in the classroom and working with middle school students. I think it's very authentic. And I think you are the voice of the classroom teacher. And we really appreciate all the work that you've done. Track 1: [1:22] You're going to make me blush. Good thing it's just a podcast. I still, every time I think, I can't even talk anymore. You've got me all flustered. You know, every time someone says that to me, I still have that imposter syndrome. I never think that people listen to our shows. So thank you. Before I go in, because I'll start on a tangent, I want to stick to the rules here. I want to start with our wins of the week because I think it's a happy win being a teacher in the summer, first and foremost. But this can be anything at home, at work, wherever, outside. With me, it's always on a soccer field. But I'm going to start with our guests like we always do. And let me get started. Track 2: [2:14] So I actually have two wins of the week. Track 1: [2:17] Oh, good. You can have one for me. Track 2: [2:20] And the first one I'm going to choose is just today. My open letter from the New York Department of Taxation has finally been resolved. Track 1: [2:34] Okay. Track 2: [2:34] After two years of back and forth, we have finally come to an agreement. That's good. And I have paid what I owe the state of New York in taxes, even though I do not reside in New York, nor have I been to New York in 10 years. Track 1: [2:53] Do you get taxed in New York even if you don't live there? Track 2: [2:56] If you do work in New York, it's caused the convenience of employer. Track 1: [3:04] I don't work in New York, so that's good. I hope it's not like that in other states, but I don't work in other states either, so I'm kind of stuck here in Florida. Track 2: [3:13] Yeah, so that's one win. And the other one I will share because I think you could appreciate this as well. In a few days I'm looking forward to seeing my grandson and just the thought of looking forward to seeing him is a win. Track 1: [3:29] How old is he? Track 2: [3:30] It's not quite a year yet. Track 1: [3:32] Aw, I love babies. I love boy babies, too. No offense to girls. But being a boy mom, it's so funny. I always gravitate to little boys. And my partner, he's always like, oh, my gosh. Every time you're away from kids, all you do is look at kids. I can't help it. And I've got a long way, hopefully, a long way to go before I have grandkids. Aw, congratulations. Good to see you. Track 2: [3:55] Thank you. Track 1: [3:56] Thank you. It's always nice. Yes. Well, my, you know what? I do have a win. So I don't know if I've said this on the show, but I'm sort of like, but we're building like, a tiny, and I want to say a house, I don't like to say house, like a shack, a tiny shack in North Carolina, just because we love it up in North Carolina, everything about North Carolina. And I built a fireplace last week. Track 2: [4:23] Oh, that is a win. Track 1: [4:24] Yeah, it is a win. And you know what? I say it's a tiny shack, but we have really tall ceilings because we were going to have a bigger shack. But, you know, it's very expensive building a house. So we do a lot of the work ourselves. And my partner, he did all the cutting and I put all the concrete block. We built out the framed it and I put all the concrete block. I was like, oh, my gosh, I hope I put enough screws in to hold up all the tiles. I have learned so much. I know how to, you know, put flooring down, move gravel, unsuccessfully weatherproof a house in the basement. So in case I ever lose my job or the robots come and take us away, I can go live and build houses in North Carolina. Track 2: [5:10] Well, you should use your podcasting experience and record you and your partner together. Building this new place. Track 1: [5:22] Where were you two years ago i couldn't use that and send it. Track 2: [5:25] In as a pilot for uh the home the home uh what's what network is that called the. Track 1: [5:31] Fixer improvement yeah there you go. Track 2: [5:35] I don't know what it's called but fixer upper and all those other kinds of shows like. Track 1: [5:38] That you know not to digress because i have a feeling it's gonna be and we're gonna come back to the show in a second but i feel like the house and building is a cursed job right we've had bets I've had, you know, have you ever heard of sprickets? Track 2: [5:51] Yes. Track 1: [5:52] Okay, we had sprickets, and those things attack if they get scared. Track 2: [5:55] Yes. Yes. Track 1: [5:58] We've had water in the basement already, and, you know, in the weeds, every time we go up there, they're like five feet tall, and there's bugs everywhere. Just saying. I'm selling the place, aren't I? Track 2: [6:11] And North Carolina, too. Track 1: [6:12] Thank you so much. Track 2: [6:13] The Chamber of Commerce really thanks you for that. Track 1: [6:16] I love North Carolina in fall, winter, and spring, but come summer and the heat index last week was ridiculous. Track 2: [6:23] It's the humidity that is very difficult to deal with. Track 1: [6:27] Well, we'll have to do a show about North Carolina, but I digress. We haven't seen each other two years ago. I was looking at the webinar that we did. It was called, Will Generative AI Make Computer Science Education Obsolete? And I couldn't believe it's almost two years, August. Track 2: [6:46] Wow. Track 1: [6:48] And I probably told Mahmoud before this show not talking about AI, but answer that question now two years later. Will it make it obsolete? Track 2: [7:00] It hasn't done it yet. Track 1: [7:05] Yeah. Track 2: [7:06] And just some background about how I ended up in this position talking to you is because I listened to the Is Hello World Dead episode. And that had my mind going in circles for days. And I just reached out and I was like, we just need to talk about this. Because in my experience teaching some introductory data science courses that use R in Python, I sometimes grapple with, am I actually teaching R in Python? What am I teaching people to do? Am I trying to get them to think like a data scientist? Am I trying to get them to become curious? Or am I actually teaching them code and how useful is the time that I spend going over code to them right now and in the future? Track 1: [7:56] I know. It's a lot to, it's crazy. It's a lot to think about. And I'm going to tell you a story about a girl that I work with. But before, I just want to do a proper introduction for you. Because this is what happens to me when I'm by myself. I don't keep control of the situation. And then I get really excited and I just want to talk like we're old friends. But tell us about what you do, data, R, Python, everything. Just spew it out. Track 2: [8:26] So my full-time job is I'm the instructional design director at a national nonprofit named Data Science for Everyone. It is located currently on the campus of the University of Chicago. It was started by one of the authors of Freakonomics, Steve Levitt, and we were founded to address an issue around mathematics and how students were learning it, in separation from context and authenticity and modern tools and data. And our mission is to make data science education available for everyone, especially students who are in mathematics. We want data science to be counted as a fourth year mathematics course. And we also work with other groups to try to figure out where does data naturally show up in their subject area and help them understand what data literacy and data science are about and have them incorporate. With the hopes of raising the engagement level of students in their coursework so that they can see a connection between the things that they learn and the society that they live in, and the futures that they might, end up shaping when they finish high school. Track 1: [9:52] And I love everything about data science. There's two stories I want to connect to that, because I think it's so important. And for my answer to my question that I asked you, will generative AI make computer science obsolete? Like, for me, I still think absolutely no. First story. So Python, show someone matplotlib, you know, you get, you just copy it, have them change the data. It's like, they don't even really need to read matplotlib because it reads itself and they get a chart. And because it's all they have to do is change the data within the X list or the Y list, you know, they can easily make a, you know, a bar graph or a line chart depending on what they do. And when they make it for some reason, it still wows them. Even though AI can generate a graph that's so much prettier, the fact that they wrote like six lines of code and it just popped a graph out, I think, that to me says a lot about the beauty of visualizations and the fact that once they then get into AI and they do some, pie plots or they use some other library in there and it makes a really cool graph, then they're like, this is cool. Track 1: [11:04] That story. And then I work with a girl who just graduated with a data science degree. And I took, you know, I'd struggled through a Georgia Tech bootcamp, six months of data science, three nights a week until 10 o'clock at night. And I loved it. And I hated it. It was machine learning, data science, Python, you name it, smushed into six months, the people that succeeded and got jobs, they deserved it because they were rock stars. It was hard because for me, the storytelling behind the data and being able to see the patterns from the data, even with AI, because I kind of did it when AI was around, it just didn't show up as nice. And I worked with this girl who just graduated and she, when we started working on this project, we have this special project at school. She was she would come to me she's like oh I did this and I saw this and I was like oh my god I didn't even see that and you looked at like 5,000 points of data and that's the pattern you got and I think that that right there is one of the reasons why if you can get someone to learn data science, the brilliance just shines through that and is so that's my story for you of Track 1: [12:23] my two stories of why it's amazing. Track 2: [12:25] Well, I want to sort of go back to your story about the visualization. So when I'm teaching students how to visualize in Python, I'm going to go back to your story, When I first started teaching, I wanted them to make the same graphs that I was making with my data so that they would learn how to do it. And maybe two years ago, I had this epiphany. I was like, well, let them choose their own data and make their own graphs. And they learn so much more when they see that they've chosen the wrong kind of graph to visualize a certain type of data. Or when they look at their graph and they're wondering, how come my line chart doesn't look like a line chart? What did I do that was wrong? And it motivates them in a different way because now they're invested in solving their own problem. And what I try to highlight to them in that moment is that you now are motivated to learn more about Map Plotlib because now you're curious. The curiosity is going to drive your motivation for learning. Track 1: [13:41] 100%. Track 2: [13:42] Start with the curiosity. Start being a questioner. Start just wanting to know more. And the more curious you become, the more advanced you will become as a programmer because now you have an internal fire lit inside of you. To find out is there an answer to my question and so i really appreciate you bringing that, sort of idea letting students play around with code and see things and how you confirm for me that that sort of lights in them this spark that is really hard to do otherwise. Track 1: [14:19] 100 and i think that's always been like a, Ask any good educator, and they know, like, if you can have them make a connection, a story of something, it's going to bring about the learning and just want people to dig in. And I do say if you can't find that, the best thing, the second best thing to do is if you have middle schoolers or ninth grader, get the boys to do a workout and let the girls just sit back. Because the boys will count how many pushups they can do, how far they can jump. And the girls are just sitting there heckling them because they think they're all like manly. And the boys are like, oh, I can do 25. And then graph that out. And it's quite fun. And I did that. We did that across two classes. And then we kind of had a competition to see who was stronger. And then we graphed like they wanted to graph get grades. And I was like, no, we're not going to do that. But anything that caused the competition helped. Track 1: [15:23] But yeah, it's funny. I wanted to ask because I don't know, I saw this in the research and I don't know if it's still saying what is this adapt model? Track 2: [15:32] So the ADAPT model is something that was developed at the North Carolina State Data Science and AI Academy. Way back, I like to tell this story that, when they started the Data Science Academy at that time at NC State, there was a postdoc student who was going to teach the Introduction to R and Python for Data Science course. This individual ended up getting a full-time job and could not teach the course. I was at that time working at the North Carolina School of Science and Mathematics, which is located in Durham, about 25 minutes from the campus of NC State. The director at that time of the Data Science and AI Academy named Ray Levy. She was a graduate of the high school that I was teaching at. So she called the high school and said, is there anybody that could teach this course? And they were like, oh, yeah, Mahmoud could do it. And so I'd like to say they were desperate and I happened to be available. Track 2: [16:37] And I started teaching an introductory data science course with R and Python. And at that time, they were developing this model because their theory is, we want all students across the campus to experience a taste of data science, in a variety of ways in a project-based course. And if they like it, then they can do more. And if they don't like it, it was a fun experience for them, hopefully. Track 2: [17:09] And the ADAPT model has three parts. And really the two most important to me are the project-based learning and the 10 common learning elements. So the 10 common learning elements, not every course has to address all of them. But what they've done with those elements, it gives the full spectrum of what, data science and AI could be. Like, do you demonstrate the mindset of a data scientist? Are you a good questioner? They're not all technical learning objectives. And the project-based learning, we were able to develop, a framework for doing project-based learning with data science in particular, whereby week two or three students have selected a data set that they find interesting. And the story I was telling you about me doing a visualization, the students doing the same one. Well, when we are learning things about coding, I'm using data. That's my data, but they're using their own data from like week four. Track 2: [18:20] At each step of the way, they learn more about coding, they learn more about their data, they document it in this reproducible workflow, and at the very end, they produce this report. And the reports and the presentations are amazing because they've been working with the same data set for like 10 weeks. And what we have discovered with that and what I think students learn from the experience is when you're teaching things with data science and even with coding, the earlier you get them working with their own data and they find something that they want to learn, but we're not going to cover it because this is an introductory course. They're like, well, what's that library? How do I use it? Track 2: [19:09] How can I do this? And I say, read this documentation. Well, I used to say read this documentation, but now maybe they use generative AR or other tools. But they're now motivated to learn beyond the syllabus of the course because they have time. And if I give any recommendation, if I were to give one recommendation to any teacher out there who's thinking of Python and using it for students to do projects based with data, get the students working with their own data. as soon as possible. And throughout the remainder of the course, they're going to have questions about things that they want to do that are not part of what you're going to cover, but they'll be motivated to learn it. Track 1: [19:52] 100%. I was thinking about this because I know you're a math person, right? Track 2: [19:59] We're all math people. Track 1: [20:01] You're a math person. You're one of those people. I mean, I say this story all the time. I think I took Calc 2, maybe. I don't know. I took so many math classes in college because I was a pre-med, and all they did was like shove me through math classes. But the one class that I didn't take that I wish I had was stats. Because I feel if I had had stats, I probably would have seen data better because no offense to math people, but Cal just didn't really help with data science. I don't really, I mean, maybe if you're building a computer and all that stuff, you need to have some sort of higher level math, but it was beyond me. And I feel, I feel like when you have a, um, You know, you have a connection to the data, but when you also have an understanding of numbers, you can get in a little bit deeper with data science. Like, what's your take for people who, okay, maybe they're coders like me, but not necessarily math lovers. What's your take on that? Track 2: [21:07] I don't think that you need a lot of math. Track 1: [21:10] Just like coding, you don't really need a lot of math in coding. But it helps because I think you're smarter. Track 2: [21:17] Well, it really depends on what you're doing. With my students, I want them to understand, before we talk about technical mathematics language, especially when I was teaching calculus, we're not going to use the mathematics terminology. We're going to speak to each other like we're hanging out at a cookout in North Carolina. But over time, if I want to explain to you what I'm doing, I could say four sentences or I could say one word. And you learn that this word means what those four sentences used to say. And so with your question about a non-math person or someone coming from computer science looking at data, what I would say to those people is don't think about. Track 2: [22:09] Data science as an academic endeavor, think about it as a process where you're trying to get any insight or gain some understanding. And sometimes that might be a graph that you have to look at. Sometimes that might be some frequency tables that you have to make. You'd be amazed at what you can learn just by looking at basic summary statistics. And once you sort of get an idea of what it is that you think you might see, possibly there might be more advanced math that could help you go a little bit further. But don't start with the idea that you want to learn the advanced math first. Start with the idea that you want to find, is there any nugget of wisdom I can gain from doing some basic computing things and using some of the basic libraries for data science that are in Python? Like I tell students all the time, I'm not teaching you to be a data scientist. I'm teaching you how to use pandas so that you can come up with really good questions and build curiosity so that you can learn more about pandas and then maybe learn more about other libraries so that you can then do more, in-depth things in data science. Track 1: [23:27] But that's the thing right getting kids to be curious i think i think that's the beauty of it i think that was when we had that webinar i was like oh yeah this guy's pretty cool we mesh, we mesh a lot i think we have a lot of same kind of philosophies of that getting kids to be curious getting them to want to learn then the learning's easy like, I feel if more, and I'm probably teachers are going to hate me for saying this, but I feel like if more teachers would just back off with the, the, the pushing and try to, you know, build the, the digging, like, what is it? Think about all those good teachers you had growing up. It's the, it's the teachers that would kind of lead you, you know, they never took you straight to the, to the, the directive, they kind of led you around this path and you're waiting there and you're waiting there and they're like, Oh, the bell rang. Bye. And you're just like, wait, wait, wait, wait, we've got to finish this story. And I think that that's what drives the kids, um, to code, to do data science, to do math. It's that, what can I, what can I do with this? What can I solve? And I think that's beautiful. Track 2: [24:43] From just based on what you're saying, like there are two people that I've done work with in the past who I think are masterful at that technique. And that's Dash Young Saver from Sku the Script and G Sun from Course Kata. They have this ability to design instructions so that you're going down this path and you as a student, you think you already know what's going to happen. And when you get to that point and it doesn't work out the way you thought, you're like, wait, what happened? Well, you got to do some more coding. You got to learn some more stats. You got to learn some more math because what you thought wasn't what it actually is. And they're like, well, can you just tell me? Well, when we learn more math, we'll come back to it. Or when we learn more coding, we'll come back to it. And so they have this ability to create this consistent yearning to learn more so that they can work out what happened later. Track 2: [25:43] In contrast to what they thought was going to happen. And so I think that's a big part of education is creating those scenarios where students are forced to sort of grapple with what they thought and what actually happened. And then they understand that this subject matter, mathematics or coding, is what's going to unlock the answer that you're looking for. And so now they have a reason to learn it. And now you're not always disconnected from real life and the content area. And that's not to say that there are some things in math or coding that you just need to learn how to do because those are the basic things you need to know how to do. But when it comes to being in a classroom with a student for a semester, if you're in a college or for an academic year, there have to be points in that year where students are grappling with things that they thought. And your content area, knowledge, is the only thing that's going to unlock the door for them. And so they're waiting on pins and needles to get to it so that they can open the door to see what's behind it. Track 1: [26:55] Well, that's pretty. That's really beautiful. Yeah. Track 1: [26:58] I have two questions that kept going through my mind when you were talking. Are you a calculator person or a no calculator person? Track 2: [27:07] I am definitely a calculator person. Track 1: [27:10] Thank goodness. See? Track 2: [27:11] All day. I'm an all-day calculator person. Track 1: [27:14] I think calculators should have always been in there. And then the second question, because this is why I love doing the podcast. It's all about me because I am just like question, question, question. I always have these questions going through my head. But why, and I know there's like some sort of historical, you're going to have to teach me this now, even though this is a Python podcast. There's always R in Python and data science. And some people just only stick to R. In fact, this girl that came to work with us, she knows R, but she didn't know Python. And I, you know, I showed her like two weeks of Python and now she's a superstar. But why R? Why was it a data science thing? and then YPython, tell me this history that I don't know anything about. Track 2: [28:00] All right. So I'm going to try to do this from the top. Track 1: [28:02] In a short version because we don't have that long of a message. Track 2: [28:05] Yeah, I'm going to try to do this off the top of my head. Track 1: [28:08] Okay. Track 2: [28:09] So one of my favorite professors ever is Justin Post, and he's a professor in the statistics department at NC State. And he would say that R was designed by statisticians for statisticians. So when you look at people who do a lot of statistical analysis and they work with data, many of those people are going to use R. The output from certain statistical modeling techniques is put in a way that they understand exactly what it means. Track 2: [28:46] In Python, it was designed to make coding more accessible to the masses of people. But once people found out how easy it was to code in Python, then just as everything else, curiosity and ingenuity of human beings said, well, why can't I make it do this? Why can't I make it do that? And so then you start getting these libraries that are developed. And once NumPy was developed, then on top of that, you had pandas. Now we have to visualize. Now we need to do machine learning. And so you get this suite of tools that you can use for virtually anything. But in R, straight out of the box, with nothing else added, you have all the functionality you need to work with most data sets, most data type, most file types. Because it was designed for that purpose. And with my students, we learned how to use RStudio, which is a common interface for people who use R. You can use Python with RStudio, but most people who are R people use RStudio strictly for R. And then for Python, we typically use Jupyter Notebooks. Track 2: [30:09] And so they get a chance to see both interfaces. I would say that our studio, is a lot more powerful in and of itself. You can do many more things with it than you can do with just a Jupiter notebook. But for education purposes... Track 2: [30:28] There is a new tool called Jupyter Everywhere. Have you ever heard of Jupyter Everywhere? Track 1: [30:33] I have not because I'm a co-lab kind of girl. Track 2: [30:37] Okay, so imagine if you had the ability to not have to authenticate in Google. You can share, just like you do with a co-lab notebook and everything runs in your browser. And that's what Jupyter Everywhere is. And it was developed as a tool to get around the idea that some districts block, Google Colab because of the authentication and data privacy. And you can set up a Jupyter lab or Jupyter hub, but you can't ask teachers to do that. You don't want to install this on everybody's Chromebook because you need developer access. And so this is a tool where all you need to do is go to JupyterEverywhere.org. You can start coding in R or Python right on a Jupyter notebook. You can share using permalinks. And it's free. Track 1: [31:42] I love it. It's very cute. Who is it made by? What's up with the octopus? You know, you can't tell me something. I'm going to go straight there. I'm like multitasking. Track 2: [31:54] So the color scheme in octopus, I had no decision-making authority in that. But it was done in collaboration with people from Skew the Script and Course Contest. And they got some funding and they developed this tool to remove barriers because at Data Science for Everyone, we're all about trying to remove as many barriers for entry into data science education as possible. And we feel like that this one is going to be a game changer. Track 1: [32:22] That is pretty cool. And so you can just run regular. So the sixth grade, I'm sorry, now you've got my head like going blah, blah, blah, blah, blah, blah. Well, I think you had told me about this before, and I was just like, so much going on in our life in 2024. I think just 24 and 25 went out the door. Track 2: [32:41] I may have told you about it before, and then it was still like an idea that was being developed. But now a lot of testing is being done with it. And as a matter of fact, we ran a test this academic year with some teachers in AP stats and AP computer science who use Jupyter Everywhere to do a little data science project in each course that we co-developed with people from SKU the Script. And some people from Computer Science Teachers Association. And we're going to see how it goes. And if it goes well and everything works, we're going to release it on the planet. And you can have middle graders writing Python code in a Jupyter notebook, saving it to a PDF. Track 2: [33:34] I'm calling this the 21st century interface for the term paper. The term paper must change. When I was doing a term paper, you get your note cards, then you do your outline, then you do your rough draft, then you do your final draft. We want to start teaching people this idea of reproducibility early. So now you use your Jupyter notebook, you turn in your first one, which is your rough draft, you get feedback, you go back and iterate, make it better, then you take it, you extend it, And all the way you have a documented workflow and students can see how their understanding has changed and grown. And then you ask them to do a reflection about what did you learn? What did you change? What would you do differently next time? And these are the skills that. That we want students to have, but we don't always have the right interface for them to do it. And I think using Jupyter Notebooks is one way that we can get to that point. Track 1: [34:43] I love it. I just, you know. Track 2: [34:46] And you can do R now. Now you have a reason. Track 1: [34:49] I don't even know if I have a reason to that. I'm sure you listened to the story about the kid who tried to teach me C, and I think we lasted like four months together. And he got board and now he did lua and now he's doing something else he was like yeah that challenge challenge done i don't want to teach her c anymore um but you can import um, matplotlib and all the libraries into oh oh he's so pretty okay well we definitely have to check that out i was just thinking how cute is it i i'm always a person that, likes to bring in younger people and definitely always trying to get the girls to come in and it's got a nice interface it's it's really nice. Track 2: [35:29] A thing that I'm very big on, and my colleague Hannah Kurzweil, she does this a lot, advocating for teachers. We don't want to design something and say, here, teacher, look at what we have. We want to do it in collaboration with teachers who are in the classroom. And so we get lots of input and feedback on what will make it better, will make it more useful, or make it more engaging. And so working through that mindset, you get these fun colors, these big buttons, limited amount of things that you can do. But they're all the things that you would want anyone in grade six to 12 to be able to do. And removing the clutter to make things clear so that the instruction and the learning objective is the thing that students are using their cognitive energy to decode. And they're not clicking around on 20 buttons in some interface to try to figure out how to make things run. Track 1: [36:28] I love it. What's some simple R code? Track 2: [36:31] Some simple R code? Track 1: [36:32] Give me like a print hello world in R. Track 2: [36:36] It's the same thing you do, print hello world. That's it? Track 1: [36:39] That's it. It's the same? Track 2: [36:41] It's the same. Track 1: [36:42] Boring. Why do they just call it Python then? What's a loop? Track 2: [36:48] So in R, you can do the word for. And so then you need parentheses and say one. No, do this. Do for in parentheses and do I in. And then do one colon 10, close the parenthesis, then do a curly brace. Track 1: [37:17] You're making you code from memory. I love this. When's the last time you coded without AI? That's okay. We'll get you there. Everybody's cringing now. Yeah, you have to do something after it. Track 2: [37:33] And yeah, then between the first curly brace, then you can put a print statement like print I or something and then close that curly brace and see if that works. Track 1: [37:46] No, I missed something. That's cool. I'll play Rob that and someone will be, doing it. You're going to be like, oh, that was what I was. I put you in the spot. Track 2: [37:53] No, no, you're fine. Listen, I don't mind being put on the spot. Track 2: [37:59] I think it's something that all teachers at a certain point get comfortable with. Track 1: [38:04] Yeah. Hey, I'm going to put you on the spot because I've been noodling and I can't believe I say noodling now, but it's actually my word. It didn't come from AI. But you know how AI always says, well, I have been thinking about this question. I'm like, no, I haven't. I've been noodling it. Stop trying to change my words. I am trying to write this stupid thing, like article I write articles and everybody probably thinks I write articles for them but I really write it for me because I'm trying to process what's going on in my head and make it slow down, so I had this thing about capabilities and this kind of works good with data science and this is why I'm bringing it up, putting AI out of the way I don't even want to talk about AI because I think this is coming in whether it's AI not an AI but in the end for the real world, whatever you're going to do in the future. You know how we have frameworks and skills and teachers have like certain check marks and we have to do something and everyone's like, oh, we need critical thinking. We need problem solving. We need this. And all that stuff, all that stuff takes, you know, our capabilities. It's like what we either initially have in ourselves and what we can develop more. Within ourselves. And one of the words that I really hate that's been floating around all over everywhere is judgment. Track 1: [39:31] I don't like that word because, and it's a good word, not that I don't like it because it's used so much, but I don't like it because people don't really define it. But if in data science, right, you really have to have a lot of judgment. You have to be able to weigh the knowledge, the context, the uncertainties, any values or consequences. And you have to be able to make it explainable, thoughtful decision on what you either see or don't see. And so I've been trying to noodle with this word to the fact that nobody really defines judgment in all the AI frameworks. But it's big. It's huge. And I think, like after talking to you, I really do think things like data science and coding and where you have these crazy amount of numbers can really help kids, develop judgment, which later then we'll be able to tell what's right or wrong, ask the right questions. So what's your thought on that beautiful word of judgment? Track 2: [40:43] When you said judgment, the first word that popped into my mind was the word nuance. And if anyone's ever taken a class with me, they probably say, Mr. Harding says nuance all the time. Like to me, everything is about nuance. And I tell my students, just because you can write code to do something, that doesn't mean you should write it. And just because you get output, that doesn't mean you should use it. And I think the judgment piece comes into play with data science when we talk about domain-specific knowledge. Track 2: [41:25] One can make a decision based on output that's been produced from code, but without understanding the domain, they have a very limited ability for it to be used in a way that's beneficial. And I experience this on a regular basis because when you let students choose data sets, sometimes they, many times, they choose things that you have no background knowledge of. One student chose a data set on cockroaches, like I don't know anything about, or snakes, or one person wanted to look at like all the different types of trails in the parks in North Carolina, things that I don't know anything about. And I love it when they do that because my first question to them is, what do you know about this topic? Track 2: [42:29] And if they say, well, not much, I'm like, you have a great opportunity to develop your ability to have nuanced discussion about this topic, because you need to learn more about it before you can work with data around it. And a typical thing that happens is some students who have NBA league paths will choose a data set about NBA players, and they want to do, a model on who's the best player. Track 2: [43:01] And I say, what do you mean by best? And they're like, you know, the best. How do you judge between one player or the other? What is your metric? How do you define it? How do you evaluate? Does age count? Does size count? Does amount of contract count? And all of these kinds of questions, they are motivated to kind of find answers to because they're working with data around it. But what it also does for students, it also refines that judgment that you're talking about. Like, I think I know what it means to be the best. I've done some work. Now I really have to look at what I've done, and what the output is telling me and use some human judgment around, is that correct? Track 2: [44:04] So when I think about opportunities that students have to do that in school, especially when I was coming through school, those are very limited because some of the math problems in the back of the book are very contrived. Like I say it all the time, like nobody cares about when two trains are going to cross on the train tracks anymore. Nobody cares about that. We don't care. Track 1: [44:27] Especially if it's a bright line in South Florida. Track 2: [44:32] We don't care anymore. But there are things that students might care about. And those things may not have definitive answers, but you need to make a judgment about what you're going to do. Like, there are no perfect solutions. There are solutions that are being worked on to try to make them better. But that takes a lot of experience. That takes a lot of background knowledge. That takes skills in many different areas, like you said, those capabilities. And I would challenge is maybe not the right word, but encourage any teacher across any subject really to try to take some kind of data, and have students think about what it means in relation to the subject area knowledge that they're learning and what the data is trying to say to them or what they think the data is saying. And have the students talk about it, debate about it, make judgments about whether or not they believe it's reasonable or not. Track 1: [45:35] You know, again, this is why I love these podcasts. I think people don't really realize it, but they're all doing a form of data science, right? Think about a book, English class, right? They give you three books to read. You take in all this information. This is a book of themes about friendships. And we're trying to connect these clues and these pieces of data artifacts from each book to try to find a pattern of why these three books are connected. And I think sometimes it's nuanced, right? Track 1: [46:12] Sometimes they pick books that are really contrived and it's like an easy essay, but the ones that aren't really connected, the ones that are so... Track 1: [46:22] I don't want to say disproportionate, I can't think of the word. Sorry, this is getting late. This always happens around 8 o'clock. My brain starts turning to mush. But things that are so disconnected and try to find the connections, that is data science. Think about history. I was trying to figure out why a teacher, and I've told this story before, this teacher gave my son silk roads. I don't know. They did the slave trade. They did three totally different points of times of history and they taught them at the same time. I'm like, what is this guy doing? Why are we talking about Asia? And then we're talking about Africa. And then we're talking about United States. I was like, what's going on? But it was about this theme of kind of like capitalism, globalization, blah, blah, blah. I'm not a history person. But I was like, that's like huge data points. And I actually put that into AI for it to help me understand the pattern. But I think if more people did what you just said, and realize that everything that we're doing and everything literally is data, like mind-blowing take the take the things that you guys do for data science for everyone and turn it into your curriculum and make them not have contrived problems where kids could actually apply judgment that doesn't have a right or wrong, i don't think people would like that because that's going to be hard for them to grade, but get over it. Track 2: [47:44] No, but it's going to make your job more fun. Track 1: [47:47] It is. Track 2: [47:49] And when you said like books as data, you can think of text as unstructured data. And it made me think of a website called Plotting Plots. Oh, see? You have to go to Plotting Plots. You have to go to that. One of the visualizations on that website is all the lyrics from the play Hamilton. And who said it, and when they said it within the play. Track 1: [48:18] Wow. Track 2: [48:19] Right. The second thing it made me think about is another website called the Pudding.Cool. Like pudding? Like pudding? Yeah, like puddingedible.cool. There's a visualization on that site that is the coolest visualization I think I've ever seen. They took samples from notes that were taken in Congress when people mentioned democracy and determined whether the statement was in, support of a strong democracy or if democracy was under threat. Track 2: [49:01] And made this visualization. If you teach U.S. history and you spend two class periods with your students, having them try to understand what's happening and break it down, it will be some of the most engaging two class periods that they've ever spent. And what you're saying is correct. When you approach the world, as a curious learner and find where it connects to your subject area. Most times that's going to be with some form of data, unstructured or structured data, or it may require data that no one has put together that your students might need to put together. And then have them use the subject matter that you're teaching them as the starting point to analyzing it, finding the nuance, making judgments, seeing if those judgments are correct, and debating the usefulness of each decision that's being made. It might be really hard to grade. But if they're engaged enough, you may not need to grade it. Just say everybody got 100. You participated. Track 1: [50:21] That's what they should do. I mean, to be honest. You participate, you're off task, mark a point off. That's what I would say. Like, they're learning. Wouldn't that be fun? I mean. Track 2: [50:33] I think people sometimes misunderstand me. And I just want to clarify, I'm not saying every class is going to be like that. It's not going to be like that. Track 1: [50:40] Why not? It's so much easier. Track 2: [50:43] But it can't be that no class is like that, right? They have to experience that. Track 2: [50:54] To do that, what it requires you to do is just let go a little bit, let students have a little bit of agency, and be okay with them seeing things in that discovery that you didn't see. When my students are trying to do things with Python that I haven't done, using libraries I haven't used, if they're willing to put in the cognitive workload. Track 2: [51:19] I'll sit there with them and try to help them figure it out. If they're thinking about analyzing data that may be a little beyond their reach, then we'll work together to see how we can code it up. We'll do some pseudo code to see if that's what is going to give us the result that we want. Then we'll test it out. And if it doesn't work, we'll go back to the drawing board. I've done it several times where I make these Google Docs with students and they're like these running notes of things to try. And then they're like, what should I do with the doc? I'm like, keep it and show it when you go apply for a scholarship or apply for a summer program to show them that you're a thinker. You're someone who is motivated by curiosity, who is willing to go beyond what the course has stated as the minimum objectives. And when you show them this, that's different than you saying, no, I'm a self-learning. I work hard. Everybody says that. What's your proof? Track 2: [52:24] And you can talk about it because you've experienced it. And I just think that this, you know, this and the fact I know we said we weren't going to talk about it, but, the fact that the fact that for very fundamental things, generative AI can give students code for students who aren't going to be programmers. Having access to that tool to do fundamental things that removes the effort of learning to code but can amplify the analysis and the curiosity i think that's a win. Track 1: [53:01] Me too i i agree i i i think like for me and this is what i've been trying to i was supposed to write this article and release it i think today yesterday i don't remember what day i'm summertime i didn't even know what day it was. I asked my son what day it was. He's like, it's Wednesday. I'm like, oh crap, I have a recording today. No. But yeah, I think AI is, can with the right guidance and judgment and curiosity can help develop that subject knowledge and bring out stuff and the foundational knowledge, to help you make decisions on non-contrived issues like if you can develop non-contrived problems and have the kids actually go in and and dig in with an ai and justify stuff but not with just saying oh i used ai to do this, that's not justifying that's kind of what these ai literacy frameworks kind of get stuck in because it's so hard to do and so hard to grade but i'm not going to go there i want to share because we have a little bit limited time, i try to cut these off on an hour and we're pushing it this is what happens when we have great conversations yeah but did you see the released um you'll i don't know if you've seen this and you might like it but the grace hopper released uh um talks, from the military 1982. Track 1: [54:28] I'll send it to you in a chat. Track 2: [54:28] Oh, send it to me, please. Track 1: [54:31] 1982, right? That's like pre-internet for the rest of the world, but military had it. She was like, we're collecting data. We're collecting data everywhere. And we're not using this data. And what is wrong with the fact that we're not using this data? Do you understand how important this data is? Talk about a woman who is impressive. And I'm watching a 1982 little spunky older woman i don't know how she old she was she was like 80 or something um, yeah it's a great one you've got you've got to send me that because they released it when did they release it oh you got to say they released it in the same month that we had a recording where that's why we didn't watch it, but i check it out because i think it's it goes into the fact that, People knew, oh, wait, did I mess up? Taking a picture of myself. Hi. Track 2: [55:23] Okay. Track 1: [55:25] No, I pushed picture instead of chats. I'll send it to you in a second. But yeah, what a great thing that she knew about the data and we were collecting all this data and how important data was. And I'm going to make this wrap it up and I'll let you end it. But how important data was even in 1982, prior to all the collections of big data. How crazy. Track 2: [55:46] Wow. Yeah, send me that Because when I was teaching my classes, I would always end with the story of someone that people didn't know who made a significant contribution to the field. And I try to switch it up so it's not always the same person. Track 1: [56:03] But she really didn't. She didn't really find the moss. Someone else found it. But she said, oh, I've got a bug in there. Or someone said it and she got coined with it. So I did a lot of research on her because I think she's a spectacular woman. Track 1: [56:18] But yeah, a little bit, four minutes to keep under an hour. Where can we reach you if people want to find out more? Because you've just shared three amazing websites, and I'm sure you have about 300 more amazing places for people. Track 2: [56:32] So if anyone wants to get in contact with me, you can just go to datascienceforeveryone.org. You can go to the About section. You'll find me. I'm on LinkedIn. And my email address is mharding at uchicago.edu. One thing that I'm very fortunate to do because I'm not full time in the classroom anymore is I take it as a, serious responsibility to be aware of as many things as possible that can help, educators, especially when it relates to data science and mathematics. So I feel like I have this like running Google Doc of all this really cool stuff. And when I'm on podcasts, having conversations with people, I'd love to share those things. And that's something that I take very seriously because I know that in the classroom, you don't always have time to spend three hours in the day looking at websites and looking for cool things and having someone, who can share those things with you. Is something that I find I get it really makes me feel good if I feel like I'm giving back at a different level and so um if anybody wants to do cool data science projects just reach out I'm always there to do that. Track 1: [57:53] I love that I feel like your best friend is data just like my best friend now is python and ai that's all I do I like what are you doing today I sat there and looked at more websites so I love that I love that you were on the show I love that I successfully did this by myself with a great you're my co-host you know you did a wonderful, job well it's great when you have wonderful um, I was going to say co-host. There you go. Wonderful co-host with you. Who needs Julian and Sean? I got my mood. You're next on my fourth call list. Track 2: [58:25] So, no. So then what we're going to have to do is we're going to see if there's a way to get you, Julian and Sean, to come to Data Science Education K-12 in Atlanta, Georgia in February. Maybe we can do a live one there and talk about data and programming and Python and all that good stuff. Track 1: [58:47] Does it have anything to do with AI there? Track 2: [58:51] Everything is AI. Track 1: [58:52] I know. So maybe I could probably go there. Just send me the link. We'll put it in the show notes as well. Track 2: [58:58] Okay. Track 1: [58:58] You know, February. When in February? Track 2: [59:01] February 15th to the 17th. Track 1: [59:03] All right. Track 2: [59:04] We say Atlanta, but technically it's Buckhead. You know how it is. Track 1: [59:07] I love Buckhead. Track 2: [59:08] Yeah. Track 1: [59:09] I spent my, we won't go in there, but my 21st birthday in Buckhead. Track 2: [59:13] Yeah, we're going to stop the show. Track 1: [59:17] Long time ago. We'll stop the show here. But thanks again for coming on the show. That's another podcast. All right. For Teaching Python. I'm going to let you. So I'm going to say for Teaching Python. I'm going to say my name. You say your name just to make me not feel weird. And then I'll say signing off. How's that? Track 2: [59:36] Okay. All right. Track 1: [59:37] Okay. For Teaching Python. And this is Kelly. Track 2: [59:40] And this is Mahmoud.