Episode 56

  Sean Tibor: [00:00:18] hello and welcome to teaching Python. This is episode 56 AI and machine learning for kids with Dale lane. My name's Sean Tiber. I'm a coder who teaches.
Kelly Paredes: [00:00:27] and my name's Kelly  and I'm a teacher who coats.
Sean Tibor: [00:00:30] And as I just said, about seven seconds ago, we're joined today with Dale lane, which is exciting for me on two fronts. One because, Dale has been doing an amazing job, bringing AI and machining learning education to kids, but also because his name is Dale, which is the same name as my brother, who is also my best friend.
So it's not a very common name. And so anytime I encounter a new Dale, it makes me. pretty excited to have a conversation with them. So welcome Dale to the show. It is great to meet you.
Dale Lane: [00:00:57] thank you.
Sean Tibor: [00:00:59] Kelly, we have, been talking about this, I think for the last couple of years, you and I have both done the, ISTI machine learning course or AI course.
And I think most of the things that we've seen in education and around machine learning and AI have come about in the last few years in this recognition that. This will be a big part of the world that our students are entering into. And Dale, it's been a huge help to us to have the resources that you've created.
just even as a way to reference our own, learning and understanding and how we can teach that best to students, but also because the resources are freely given and available to any teacher who wants to use them. So it's been a, really a great gift to our community, and we're really excited to talk through everything with you today.
Dale Lane: [00:01:45] Thank you very much. 
Sean Tibor: [00:01:46] Dale, before we get into introductions about yourself and who you are and where you come from, we just wanted to start with the same thing we always start with, which is the wind of the week and will be a little bit interesting to see how people can guess. Your day job versus your hobby from the wins that you share with us.
So, we'll make you go first. Cause that's the way that we do it  when we have a guest on the show, it makes a little bit more fun and interesting for people to hear your voice instead of Kelly's in mind. Dale, is there anything that's happened to you this week that you would qualify as a big win?
Dale Lane: [00:02:14] Yeah. so jumping ahead of it, to some machine learning for kids stuff, but I'm going through a big thing at the moment of trying to move a lot of what the site can do off of backend servers and something that can run on students' own computers. And there's loads of benefits to that, but it's been a huge piece of work, but it's only finally really come together in the last week or so.
And seeing now the kinds of projects that students can do on their own computer without having to set up. accounts on online service and stuff is I'm really pleased with it. So yeah, it's a very geeky, very deepen and Gopi stuff, but that has been a huge part of the last week for me.
Sean Tibor: [00:02:52] That's really exciting. 
Kelly Paredes: [00:02:53] I was reading about that, following your blog and I can't believe how long you've been blogging, like 2011 or something. And I was reading your recent post about it,  I'm looking forward to that of seeing when you're launching it for the individual students. That'll be really cool.
Sean Tibor: [00:03:07] it's, always strikes me that there's a lot of work that goes into making things simple. So moving something on to the student's local machine and making it simple and easy for them to access is a lot more complicated than I think people realize when they're just looking at the end product.
Would you agree with that?
Dale Lane: [00:03:22] Yeah. and something that we'll work on the kinds of computers, the range of computers that are used in code clubs in schools. Cause it's, I spend a lot of time in my day job working on code. That's gonna run either on fairly powerful. Laptops or, high-end servers. and actually you sort of see that the kinds of computers that Koch clubs are using even to dance the extensive, like a raspberry pie or a pie top or something like that.
and it does. Yeah, it's a whole new thing and the whole different thing of trying to get something that will run and that smaller footprint.
Sean Tibor: [00:03:54] well, we're excited to talk more about that during the main part of the show, but I'm glad to hear that it's coming together and it's always satisfying when you've been working on something for that long and it starts to come to fruition.  we'll dig into that more in a little bit. Kelly, would you like to share your win? . 
Kelly Paredes: [00:04:07] So yeah, so my one of the week is pretty simple, but I'm very proud of it. And, In light of the sixth graders finishing up their turtle project and wanting them.
They wanted me to. Live code in front of them. And I was telling them, what do you want me to make? And they're like, make a Turkey. It's like a Turkey. So I started live coding, a little turtle project with them, and I managed to code, with functions, all the different parts of the body with the head and the feet and the tail.
All separate so that I could teach them how to do functions that would combine together. And then we can make hundreds of turkeys on the screen. And I was a little bit nervous at first because I, me, I always have something prepared just in case I mess up and it actually turned out and I'm a little bit addicted.
Now. I want to put some more like that thing on the bottom of the Turkey, and, play around with it. So that was a big, funny, silly win. I had to have a silly one. So you saw it. It,
I T yeah. 
Sean Tibor: [00:05:06] out really cute. And I think there we're right in the stage of our course right now, Dale, where we're getting into how to break bigger problems into smaller ones. And the turtle module has been fantastic for that because they can really. See it for themselves, they can visualize it and they can see, okay, this is the part that I just wrote.
This is the next part that I wrote. And sometimes that's a little hard to do with something that's console based or even a, in a notebook because so much of the functionality happens without being visualized. So it's pretty cool,  approach to teaching them.
Kelly Paredes: [00:05:38] It was fun. It was silly.
Sean Tibor: [00:05:41] And they definitely appreciate that.  for me this week, the biggest one was teaching students about the date time modules in Python.  we've gone through all the basic data types, but now we're talking about some things from the standard library that they haven't really played with before.
And we went through and did a date time, less than talking about different dates and times and time deltas. And the cool thing about it was we had it all relate to the back to the future movie. So we were finding all the dates and times in the movie and calculating the deltas between them and they really got into it.
They got excited about it. and we figured out that if you were to go back in time to the same exact date in 1955, that Marty McFly did, From here. it would be something like 1990. It would be about that same time Delta. So it was really cool for them to be able to take the calculations that they would normally do on paper or trying to calculate it in their heads, turn that into Python and then have a sense that it's working and that they can verify that it's correct.
 they were pretty excited about it and they had a lot of fun calculating all their different dates. And time Delta is like, how old am I? Exactly in days and minutes and seconds and everything.  they had a lot of fun with that.  it was a pretty cool lesson and they really got into it.
Kelly Paredes: [00:06:57] Yeah, it was pretty cool. I saw that. And then I was trying to get into the dates and put that into the birthday out too. But you know, maybe that should be my fail.
Sean Tibor: [00:07:07] Well, I, it was funny because my fail was closely tied to that, which is the Western that I did leading into this was, trying to build a class schedule. So building their schedule of classes that they have every day, and it was way too much, it was way too much typing. They were confusing so many things together and I.
Just saw that all the kids were trying to keep up with the coding that we were doing, and it just was the wrong lesson. So it was a pretty big fail. It was a waste of a, of a lesson, but we extracted some information and it helped. .
Kelly Paredes: [00:07:38] we both learned from our students this week because I'm trying to get away from having the code in front of me. I'm trying to do more of, little things, trying some new functions in daytime  looking up some more documentation on Python live without having it all.
Pretested just to show the students  how to, search for things, which is very scary and it just opens the door to all kinds of problems. And luckily I had a really smart kid, or I have a lot of really smart children  in my class and they're like, ms. This parade is, you have to do this and this, and don't, aren't you print trying to print out this variable.
And I'm like, Oh yeah, thank you. So that was a minefield this week is, just, I guess it's a fail in a win. Of just not being able to complete everything, but then having a student correct it. So what about you Dale? And he fails.
Dale Lane: [00:08:27] I can't think of any, which sounds terrible. probably means I've been to many that none of them are sticking out.
Sean Tibor: [00:08:35] Yeah. And I think that it's funny sometimes. I think failures often get erased by successes. So it might be one of those things where you had a bunch of failures in a row trying to figure out the or project of moving everything more locally. But they've been, erased by the success that you've had actually getting it to work.
Dale Lane: [00:08:51] Yeah, definitely.
Sean Tibor: [00:08:53] Dale, tell us a little bit about yourself. So, you were named, I think it was what, two years ago. sorry. Yeah. 2018 volunteer excellence award from IBM. for all of the work that you've done in the community, helping coding clubs, helping kids learn about, machine learning, AI. How the world is changing and shifting around us.
 how did you get to that point? When did your volunteer work start? what's been the focus of your work and, tell us a little bit more about yourself and how you came into that role.
Dale Lane: [00:09:19] okay.  I'm a developer for IBM. I have been for many years now, 15, 16, something like that. and. I think I started all of this stuff, the machine learning for kids stuff random. I love the 2016, 2017 sort of time. So I was working on some of IBM's AI tech for IBM projects and for our customers.
mainly, it started as a way of trying to explain to my kids what I do at work all day. so in like weekends, and particularly in that school holidays . there was a few projects that we did where  . I made something with them where they got to use some of the. I take the, I was working on at work, but using it to make a game because I sort of thought that was the best way to explain it to them because I tried to explain, this is what I do, and this is what I make.
And, they weren't really getting any of it. But when they started, I think the first thing we made actually was a guest who game. So we trained it to recognize characteristics on faces. if some, if a face has got glasses or a mustache or a hat or something like that, I'm trying to recognize the meaning of the kinds of questions you get into guests who game, do they have a hat?
Is it a man or a woman? Are they wearing glasses? over a few days we made the guess who game, where we train the computer to be able to do it. and that made sense to them suddenly they understood, what I was on about when I talked about machine learning and AI and it sort of clicked and it made sense.
so yeah, all of this stuff really started just. things I was doing with my own kids that I blogged about and, I put on YouTube and Twitter and whatever, and people heard of it and it just sort of grew organically from there. what started from, one of the other parents at a local school saying, Oh, I've, so what you're doing with your kids, do you mind coming in and showing my class?
And then I started doing that at. The local school to where I work. And then another school down the road heard and said, Oh, we've heard about the lessons you're running with them. Do you mind coming and doing it with us? and to make it easier to do that kind of stuff, I thought, well, I'll put it on the web and then I can just turn up at school and it will, there'll be able to access it.
and next thing I knew, loads of schools are using it. And so, yeah, it all sort of happened accidentally. There was no big plan. but three years later, it's now sort of become this big thing that's being used everywhere.
Kelly Paredes: [00:11:34] Yeah, that's awesome. I think, one of our former colleagues went to a raspberry PI conference in, did you go to the raspberry PI conference or, and share it, or maybe you shared it on a S on a website or something and they found it and they brought it to us two years ago. Three years ago almost.
Yeah. When it first came out and it was like, we're going to teach machine learning in your two weeks of curriculum. And I was like, okay. Wow. I'll give it a try, but how cool it would be to have a dad who's a developer who can just make a game. Oh yeah. We're going to make a machine learning game today for the weekend.
I don't know. Sean's very envious. He's 
 
Sean Tibor: [00:12:14] so Dale that, so with that kind of start where it grew from an organic. idea and grew from there, whereas it today, so for the people who are listening that may not be familiar with the tool kit that you've created. Can you describe a little bit about what's included how it works?
what kind of resources you've put together for teachers and for students?
Dale Lane: [00:12:35] sure. So, before all of this started, I was a volunteer for code club, which is, I think they're in other countries now, but certainly in the UK, it was this thing where it was, A structured way of introducing kids to coding, mainly delivered as scratch. Although, they've since broadened out to do other stuff and I'd seen that work really well.
Basically you give someone a recipe, you give someone some instructions to follow for how to make something, but give them the freedom to go off the path and change it and tweak it. But, and they have that recipe as the starting point. So everyone can, even if they just follow the instructions, they get to the end of the session and they've made something.
And, by, by doing that, they've sort of hopefully learned some bits and the more creative ones will take it in all sorts of directions. so I sort of shamelessly have copied. That pattern, just over three of the lens and, in terms of AI and machine learning. So with all of it, it's this idea of the best way for kids to learn about what AI is and what it's for and what it can do, and the implications of it in the world.
Is by doing, by making. So it's a tool online toolkit that kids can train machine learning models and loads of different types of machine learning models. So, recognizing types of texts or styles of texts or the meaning of text, recognizing pictures. And that can be, what's in a picture or the visual style of a picture.
Recognizing stuff like sound recordings, recognizing sets of numbers, all sorts of different things. So I've tried to make it as open as possible for the kinds of projects you can do. And then once you've trained the model, I've sort of come up with integrations with a few different platforms. mainly scratches, the one I use most often so that kids can use the model that they've created, to make something, unlike say the starting point is I write these worksheets that are step by step instructions for how to make  something.
and particularly when I started those worksheets were. Mainly based on things that I'd done at work. So I'd taken some real-world use of AI that I'd seen either that I've worked on myself or that I know people who've worked on it and just try to make a really simplified version, that kids can make for themselves in a lesson.
and that's been the main goal. Let's try to open kids' eyes to how AI is used all around us  to the way that it's used in the real world, by getting them to make a tiny version of it themselves.
Kelly Paredes: [00:14:52] as a new coder, when someone tells you're going to go in there and you're gonna code some AI, it's very scary at first. But when you enter into your website, it's so for lack of better words, comfortable it's it's. It's not something that's intimidating. In fact, I was showing our counselor who always tells me that she needs more tech training, but I said, Hey, let me show you real quick.
I'm going to train the computer. How to identify the difference between a cup and a face. Granted, they were totally different, but I didn't want to do it quick. And I showed her and she was like, Oh my God, that's really cool. You're teaching the computer. And I did it. And I don't know, like 15 minutes, it was something that's very easy and you can.
It opens up these conversations for people that do not necessarily understand AI or machine learning that well. So thank you for that.
Sean Tibor: [00:15:42] I noticed also that you have a lot of resources that are designed for teachers to get started. So there's lesson plans, there's guides, there's things to rehearse and practice. a lot of the prerequisite setup is very well-documented. so if there's anything that you need to create or, prepare for with the class, it's all detailed there.
where did that come from? Was that primarily upon request? Is that something that you anticipated, people knew would need to know? it's something that you don't see in a lot of other tutorial style guides or resources for teachers, at least in our experience.
Dale Lane: [00:16:16] I guess it's sort of, so I'm not an educator by any means, I'm definitely a geek, I'm a code monkey and the stuff that maybe is obvious to teach us is not at all obvious to me. So I guess I'm more inclined to write that stuff down because it's the stuff that I've had to figure out. Quite painfully, I mean, pretty locked down.
It's been a while since I've actually I've been in a classroom, but before this year, all of the worksheets, there were a few local schools that have been really generous letting me use their kids as Guinea pigs. So I would go and take like a first rough draft of a worksheet into a class and some bits of it would go horribly wrong.
Some bits I wouldn't have even thought of would go horribly wrong. And it would be, after I'd tried out a few bits and those mistakes, I tried to capture in the sheets, like basically when I did it, this went wrong. and sometimes that's technical and sometimes that's just about right how I pitch it.
the level I pitch at. I remember the first time I did a worksheet that was based on the phrase, judge, a book by its cover. This idea that if you show a machine, if you train the machine learning model with loads of different covers of books, Could a machine learning model learn to predict what kind of book it is just by seeing a picture of the cover.
and it's the kind of thing that instinctively we as people do, you recognize a thriller book or a, An action book or a cookery book or a romance book just by the use of fonts or the color palette that's used, or, the types of pictures. if there's, if it's black, if it's stars, if there's a spaceship, it's a scifi book, but what the point of the project was, well, could we say for machine learning model could start to recognize those patterns.
and I went into, what we call a primary school, And I started talking about, we were going to train a machine learning model to recognize yannaras books. And Gianna was not a word that I think any of them had ever heard of before. Like they had no concept of it cause to them what I hadn't really thought of in a school library, the books are basically divided into fiction and nonfiction at that age.
This idea of dividing it up into genres, which made total sense to me. Just, they didn't understand what I was talking about. So it's things like that I just, I got a chance to practice. I see what works and what doesn't, and because say none of this is really obvious to me of how do you explain things to young children?
I try to capture all of those, the pitfalls and the mistakes. I mean the biggest mistake I've made is just the name of this, the address of the site. the number of. Lessons I've done with six and seven year olds where they'll spend the first five minutes trying to type in machine learning for kids.co.uk with spelling mistake 10 on every attempt.
like if I knew what I was doing, I would have picked a much shorter web address.
Kelly Paredes: [00:18:51] summed up like the past three years was Sean. And I think because he's been watching me hit all the pitfalls and coding and granted, Sean doesn't make hardly any mistakes anymore, but he's a new teacher. So I would sit there in the back corner, like putting my head sometimes at a face Palm because he said some like word that I had to go Google anyways.
And it was very cute. So yes, machine learning for kids. We always bookmark that. I always put it as a link future reference. Just go in there with a Google doc that says open that's awesome.
Sean Tibor: [00:19:24] Yeah. and so since you've, launched this, I assume it's been, pretty broadly distributed. Has there been anything that's surprising to you or new and unique takes on, how to use your resources in the classroom that maybe were surprising or delightful or something that was unexpected and fun that came out of it?
Dale Lane: [00:19:45] Yeah, I definitely, I mean, it's lovely when I hear of and see schools using the worksheets that I've written, but I definitely prefer it when I see schools and clubs, making their own ideas and just using the platform as a generic tool. I've seen some really lovely ones. there was, I think it was in Canada.
I think that was a school. End of last year where they had one, a student in the class with down syndrome and who had some difficulty speaking. And they found that, although they all understood him, cause they've got to know them in that class. They found that supply teachers are new people to the class, often struggled to understand.
So they tried making a speech recognition thing that would recognize the things he said most often and basically act as an interpreter. full and they made this for their classmate for a real problem that they saw them having. And I thought that was lovely. I mean, that's the idea of not seeing it as doing AI for AI sake, but just seeing it as another tool in the toolkit and using it to solve a real world problem.
I mean, that's fantastic. So when I see projects like that, it's really lovely.
Kelly Paredes: [00:20:50] that is really nice. It's nice. when you can teach kids or give them something and code and have them turn it into something social good and, use their powers to make a positive digital footprint for, Years to come. That's amazing.
Sean Tibor: [00:21:09] that's the real breakthrough in what you've done here is that it makes it something that's an accessible tool for students at a variety of ages. And one of the things that we have struggled with early on is that. The complexity of setting up the tools or creating the code that we want to, to work with, often gets in the way of that.
Problem solving and the creativity.  when we're able to leverage the tools and the resources that you've created, it gives us the ability to get the student really in touch with the problem  that they're trying to solve and different approaches or different ways that they could choose to solve it.
that's where we have started to see some of the benefits of it. And we're continuing to focus is on this idea of, okay, here's this tool that's. relatively simple to understand and utilize with the way it's been put together. Now, how could we use this tool to solve some problems that we see in the world around us and students get really excited about it because now it's within reach for them.
It's within their grasp. Instead of having to go become an expert coder, to be able to get started on their problem.
Kelly Paredes: [00:22:17] Yup. Yup. And what I really like. Is when you open up your Python projects? just coming from I, I teach sixth grade and seventh grade. And so we just do the basics of Python. if you open up a project, I can say to them, listen, pretty much. The only thing that might be new to them is the request.
go get this website, but that's even something that they can understand. So it is something that is tangible down to, like you said, well in Python as basic learners of, so it's pretty cool. If I can understand it's a test of yak.
Sean Tibor: [00:22:59] so Dale what's next? I mean, we've talked a little bit about your current efforts to make this more local to run it, more on the device that the student is using. what do you see, coming next for this project and the work that you're doing?
Dale Lane: [00:23:14] there's always more ideas than I have time to do. so there's a few things that I've been working on this year. So a big one is around. Trying to explain what's actually happening under the covers. So there's, I mean, it's a big thing in machine learning in general, we describe it as black box problem.
This idea that we get answers, we get results. We get output from these machine learning systems, but we don't really know how they arrived at that. and. And that's the constant question I've had from teachers as well. here's a tool that you give it, the training examples, and then you use that to train a model and then you use the model to do things and you can infer a certain amount about how the tech behaves just from seeing the answers it gives.
But what people have really wanted, particularly with the older students is to know actually what's happening under the covers. So it's a really hard thing to explain. and I've sort of been making baby steps towards that this year, but there's definitely more, I want to do on that of trying to explain
without while still keeping it accessible.
What is in your own network? What is deep learning what's happening with the training examples that you're giving? Yeah, it's definitely been it's way harder than this stuff. cause it comes back to, I'm a coder, I'm not a, a teacher in terms of just writing code to train a machine learning model and use a machine learning model.
That's easy. I can do that, but explaining it so that it's clear, that's really hard. And I've spent months and months going through iteration after iteration of. Of, visualizations and explanations and wizards, and then I would sort of look at it and go, no, that makes no sense until, so, there's more I want to do on that, particularly around image models, which I haven't really started yet of how'd you explain what's happening when you're training a computer to recognize what's in a picture?
yeah, so that's the next huge big feature that I think will be coming. but there's 1,000,001 like minor sort of iterations and improvements that I need to do as well.
Kelly Paredes: [00:25:11] it brings to mind this book that I read, because I was trying to be able to explain the cloud. and it's some, one of those abstract things just like neural networks, it all happens. And I always defer to, this magic happens in the background. but there was this book called.
Explain the cloud to me like a ten-year-old. So if you haven't read that book, maybe that will help you give some inspiration on how to explain like the neural networks, because it actually helps us solidify a little bit and it was very, easy to read. So I'm looking forward to that because I still think that magic happens whenever it passes through all those, Checkpoints in the neural network.
And I walked through your tutorial, your video, which was, it was really good. So, and I like where you're going with that.
Sean Tibor: [00:25:58] Is there any, any particular sites out there other than what you've been working on that you've been using as inspiration? Like we've also looked at the Google AI labs quite a bit in terms of some of the examples and demonstrations that they give. are there any other, sources that you've seen that are useful and trying to explain some of this.
Dale Lane: [00:26:17] I really liked the stuff that apps for good do there. an organization who do, who produce computing, teaching resources for schools in the UK. and it's a not-for-profit, but I think a lot of their staff, are ex teachers. So they're coming at this from a deep set of expertise of teaching. And I became aware of them because they've used some of my resources, but they supplement it with.
A lot, all of my stuff is just, coding sessions, an exercise to build something out of computer, but they put it in such a better context, they'll, they include presentations and unplugged activities and pen and paper, activities, and discussions and all sorts of stuff that, that does help make it make more sense and not just be purely about the.
Sitting at a computer and making something. So I think that's an important part. And I definitely think you learn a lot by doing, by making something, but that, that's quite limiting. And I think I really like the way that they've, puts it in a broader context and build up to the actual practical, coding session.
Sean Tibor: [00:27:23] so if a teacher wants to get started, I mean, there's plenty of resources there, but do you have any places where they should maybe look first? and maybe we can take it by age level. So if you're working with very young kids, what's a good lesson to get started with that you would recommend to a teacher that's new to this.
Dale Lane: [00:27:39] so one that I've added fairly recently, but I haven't had a chance to. Test on a class yet, but I think would be a good fit with really young kids is, it's you train a model to looking at glass to say if it's half full or half empty, cause it's a really just sort of intuitively you can sort of predict how the model is going to pan out.
if it's less than 50 significantly, less than 50, it's going to say half empty. If it's more than 50, it's going to say how full, So it trains really quickly with a tiny number of examples and conceptually it makes sense of what it's trying to do. and it's nice and visual. so that one, I think I only wrote it a month or so ago, so I haven't really had a chance to test it on anyone, but my gut says, I feel like that should be a nice, simple intro.
Kelly Paredes: [00:28:24] want to ever test on some American kids. You're more than welcome to test. My kids love being Guinea pigs.
Dale Lane: [00:28:31] Thank you.
Kelly Paredes: [00:28:33] Sorry. I'm going to ask the personal knowledge thing. So  on the, what does Twitter think the scratch project are you? Considering putting that into a Python project.  I've just learned about it on, Python tips on how to do that  in Python, on the sentiments in Twitter.
So are you gonna make that into a project that would
Dale Lane: [00:28:52] I should do. I've got, I'm, I've been in theory, the site is scratch and up in and bison, but almost all the worksheets are scratch. Cause that's what I tend to use. I was lucky actually the app inventor integration entirely. I didn't write it. It was contributed by, some people in the U S which was very generous of them.
 it was one of the best things I did with the site was open source it and, invite contributions from anyone. originally open source within IBM. So it was accessible to other IBM employees. and I got a few contributions from people around the company, which was fantastic.
then I just put it totally open source and yeah, that happened, meant stuff I've done virtually nothing, on that's been entirely contributed by others, which was lovely. but yeah, so it does mean that I tend to be a bit fixated on scratch stuff cause that's what I'm used to, but 
I should do more in Python and app inventor and write more worksheets for it.
Kelly Paredes: [00:29:43] Well, maybe some of our followers  like to do open source things and maybe they'll get in their hands. And they have a lot of our fathers also have, or a lot of the people that we know who are Python developers have a lot of kids, hint, some people down in Australia who have lots of things.
I'm sure they would like to get into some of these. Projects with, their kids. It'll be, it's a lot of fun. I like having it cause kids like searching online for a picture and they like doing the activities I've done. The make me happy. I've done the  smart classroom. I try to do the Titanic one, but I don't know what happened.
I got lost. I think it was before I knew anything about Python. And then I did the quick train with the cup and face, and that was really simple. So
Dale Lane: [00:30:26] I find that chatbots is a good one to use, particularly when it's not a coding group. Cause I think it's, there's a bit of a different vibe when I'm going to do like an after-school club where they've opted in to doing coding after school. So they're already fairly motivated. Whereas when you're doing just a normal school lesson and it's the whole class and they haven't had the choice, those ones are interesting and.
In some ways they're the better lessons, because. Different students will get different things out of it. And I find chatbots really good for that. so in that activity, it's basically you take some topic that the class is already doing and then get them to train a computer, to answer questions on that topic, to make, something that can do a simple Q and a, and it depends.
 I'll basically ask the teacher, well, what's the most recent topic you've done. so, maybe it's something like Vikings or Romans, or I remember we did an. That was an English class I was working in and we did a William Shakespeare one. So we trained a chat bot to answer questions about, the life that biography of William Shakespeare and the place he's written.
I've, we've done ones on different types of animals before. But the nice thing is when students do that, some of them are getting into the AI and the machine learning. And they're interested to see how does the computer learn, how to recognize the intent, the meaning of a question, but some of them are more interested in the actual topic itself.
some because we always make like an animated avatar that sort of. Speaks the answers back. So the more artistic ones, so we'll get really into designing the advertisement, animating it and making it blink and its mouth move and stuff. so the nice thing is that with projects like that, they all get something out of it, even if it's not all about the coding for them.
And I do like that idea of. Making a lesson that the AI, the machine learning bit is incidental. It's, the focus of when we do the chat bot is the research topic, whatever it is they've had to go and do research about and come up with questions and answers for. And the fact that actually we're making a chat bot is almost.
We like, we try not to call that out too much in a way. So it's you get to the end and go, and by the way, you've also learned a load of machine learning and coding stuff, because you've made this interactive chat bot. so I liked doing lessons like that, where it isn't beating them over the head too much with now we're going to do AI.
and because I do think that's important. I think it can't just be about the ones who really like coding to learn about a new way of doing coding. I it's such a. An important thing for society to try and grow our literacy in this stuff. and projects like that. I like, because they're a way of trying to engage with the ones who maybe traditionally aren't as interested in AI or would be turned off by the idea of I'm going to learn about AI now.
Kelly Paredes: [00:33:04] Yeah, I can imagine Sitting in high school and having to learn the theory of AI and machine learning without, doing it the hard way where you can have this easy tool where you can get into to do the fun, get them hooked. To learning about AI and machine learning and do it the easy way speak in next week is hour of code or not is a CS week, December 3rd.
And a lot of people are doing an hour of code, and this is one of these things where kids can get in and we can host an hour of code and do scratch or do Python. So definitely we'll be excited about launching this Python podcasts during next week. 
Sean Tibor: [00:33:41] I was going to add to that, I was listening to a new podcast,  yesterday, while I was out for a walk and it was, all about, Electronics and learning or teaching electronics as part of, a curriculum. And the really fascinating thing for me was they had a, an educational design, expert on the show.
And he talked about how, when you're learning something or as you're developing knowledge, it's two tracks to train tracks that are running together side by side. And one is the track of all the things that you need to think about. as you're doing it. And the other track is all of the things that you need to do.
So when you're working with this, and I think this is where the. machine learning really flourishes was if you can give them something that they want to do, that's really strong and compelling. That is interesting for them. It helps their learning move along. It also helps them spur the thinking that they need to do to be able to make it happen.
So, that idea of. No, the, the AI and the machine learning maybe is the less prominent part of the learning and it spring all of this other work or the, all this other learning and thinking that they're doing is a really powerful way of integrating that knowledge across a lot of different areas and really engaging students' interests in their own learning.
Dale Lane: [00:35:00] That is important because in the same way that. Now maybe we don't think about learning to use a word processor or a presentation like a PowerPoint or keynote or whatever as being a lesson it's I'm right. Those are just tools that we need to do any job, not just very technical computing jobs.
AI is going that way as well. give it another five years. Maybe not even that long. will just become another tool that we all use in every discipline and every field, not just competing. It will be another way that we interact with data in the same way that, if you know the basics of Excel, you can chop slice and dice data.
And that's really useful. There's going to be, I think, an expectation that. AI and machine learning will become just another basic tool that we just use everywhere for everything. So it's trying to prepare kids for that kind of future. I think is really important.
Kelly Paredes: [00:35:48] That kind of idea that it is coming around is funny because in my newsfeed, I get a weekly newsfeed, on some top books and this is not a brand new book. It's actually a year old, but it was really funny that I got it yesterday.
And it says, it's a book called you look like a thing. I love you how artificial intelligence works and why it's making the world a weirder place. And it was this. This kind of, you need to learn about those. You need to think about this stuff. That's all happening around you. and we've got to, we've got to embrace it and be on the positive end of it, and make sure that we're understanding how AI and machine learning can make the world better.
So that was cool. Some I'm looking forward to reading that book.
Sean Tibor: [00:36:37] Yeah. So I think this is a good place to wrap up is we're looking ahead to what's next Dale. If people want to learn more about what you're working on or follow along with your progress, as you're continuing to develop the toolkit, what's the best place or best way for them to do it.
Dale Lane: [00:36:52] the ridiculously long web address I mentioned before. So machine learning for kids that code at UK, you don't need to create an account to try it. There are instructions, there have a play. and my contact details are on there as well. So I'm always, up for helping if anyone has gets stuck or has any questions, or if I can help at all.
Sean Tibor: [00:37:10] Well, we will definitely put it in the show notes, so nobody has to type it out. and that way we'll make it just a click away. for us, if you'd like to continue the conversation with Kelly and I were both on Twitter. So our show handle is. at teaching Python, I am at SM Tibor on Twitter, and I have not researched a funny social network that I created 10 years ago and still have a listing on there.
I have been trying to find my MySpace page, but I think it has finally gone away. Kelly is at Kelly Perez on Twitter. Our show website is. Teaching python.fm. you can fill out the contact form and send us a note there. We got a lovely note from someone, earlier this week that just said you make my world brighter and it just made my day brighter to get that.
So, any, conversations with us are always welcome. we look forward to, Our next conversation with you, Dale, I'd like to check in with you in, another six months or a year to see how things are going. but thank you very much for your time today. It's been a pleasure getting to talk to you about the tools that you've created and that we have used with great success in our classroom.
It has made our teaching more effective and our students more engaged. So thank you for all the hard work that you've put into it.
Dale Lane: [00:38:21] hi. Thanks very much. It's yeah, thanks for having me on. And, and it's been lovely to me.
Kelly Paredes: [00:38:26] Absolutely. And if you have need any Guinea pigs, like I said, we have a, our school is PK 12, so you can go as low or as high as you want to go. and we are always up to be the Guinea pigs. We are huge supporters of what is it when we buy things that are. Brand new and could be broken when we 
get it early adopters, whatever you want.
We are, we're love to be Guinea pigs, but thanks
Dale Lane: [00:38:51] Sounds good. 
Kelly Paredes: [00:38:52] Thanks so much for, talking with us.
Sean Tibor: [00:38:55] Yep. So for teaching Python, this is Shawn.
Kelly Paredes: [00:38:57] And this is Kelly signing off.