Podcast: AI4K12 Guidelines and Getting Students Passionate about Computer Science

We’re listening to the Getting Smart podcast where we unpack what is new and innovative in education. I’m your host Jessica and today we’re exploring the future of work and how all students can be prepared for what’s ahead. Dr. Christina Gardner-McEwn, an assistant professor in the Computer and Information

Science and Engineering Department at the University of Florida is joining today’s podcast to talk about how a love for astrology led her to computer science, the big ideas in artificial intelligence that all students should know, how we can encourage more women and minorities to learn about computer science and her predictions on the future of work. Let’s listen in to Janice and Christina’s recent conversation.

Dr. Christina Gardner-McEwn, welcome to the Getting Smart podcast. Oh, well, thank you for having me Janice. I’m excited to be here with you today. I’m excited to talk to you because we had the opportunity to talk about the future of work a few weeks ago and I’m really excited to hear more of your thoughts on what’s ahead.

So let’s start off with your background. How’d you end up choosing computer science as a path? Oh, I have like a really interesting story about how I got into computer science. Surprisingly, I started with astrology. I was really, really interested in astrology when I was high school.

So like every student in high school, I was trying to figure out who I was and what I was interested in. And so I used to spend hours on my bed with a box of coloring pencils and stacks of astrology books, charting out who I was at the time of my birth, trying to identify like what were the particular planets, what location were the planets in and what sign and what did that mean for who I am now.

And so by the time I was 16, I actually got really, really good at drawing these charts. And I was getting more and more requests from friends to draw them, but they would take me hours, entire weekends, coloring the charts, performing calculations. And if I miscalculated something, it’d throw everything off and I’d have to start all over again. So as any good scientist would do, I was triple checking things, but stuff would sometimes

be wrong. And so I found myself in a computer science class about a year later. And I was learning about programming and conditionals and the light bulb went off. Like, wow, I can actually use a computer to solve all the problems I was having with drawing these charts.

Maybe I should try to do that. And so I actually did. I embarked on creating this crazy astrology program that would do all the calculations for me and had a really large data set. To this date, I’m not actually sure if I ever finished it, but I know that it definitely

inspired me to want to become a computer scientist because it gave me the tools to solve a problem that I was passionate about. And so now you work in the Department of Computer and Information Science and Engineering at the University of Florida, where you focus on the integration of computing across the elementary and middle school curriculum.

Yes, that’s right. I am the director of the Engaging Learning Lab. I have seven PhD students and five undergraduate students working in the lab. And we work on a variety of projects that focus on studying how people learn to code and develop identities as computational thinkers and computing professionals.

And so as part of that work, what we’re really looking at is how do we design engaging experiences for students both in elementary school, middle school, high school, also undergraduate, such that they can see themselves as software developers, game developers, mobile app developers, robot designers, and cybersecurity specialists through hands-on and project-based activities.

And the goal is to allow them to develop the skills they need to bring their ideas to life, or if they’re working for a client to bring their client’s ideas to life, and that they feel ownership of computer science. It’s not this foreign thing that other people make technologies for them. It’s that they are able to make technologies for other people and to solve real problems.

So in the same way that I was inspired to become a computer scientist, I want to inspire other people to become computer scientists and give them tools, give them access to these really, really powerful tools that I think seem magical at times for people, but they’re really not. And so that’s my job is to demystify computer science and make it accessible. That’s really great. And you said two things I want to expand on, and it leads me to a story

that you told me the last time we talked together, because you spoke about really helping students bring their ideas to life and demystifying computer science and coding and making it accessible. And I love the story that you told me about a student that you worked with in your, was it your postdoctoral program? Yeah, it was in my postdoc.

Can you just tell that story real quickly? Because I think that’s a great example of what you are, what you just described. Yes. So when I was in my postdoc at Georgia Tech, I was running a summer camp program for high school, you know, basically juniors through seniors that were maybe interested in computer science. Like half of them were interested, half of them were like, my parents said I had to come.

And the program was designed to engage underrepresented minorities, so women, students of color into computing and to do it in a way that bridged their interests. And so at some point, we went out to schools and we were recruiting and I met this young lady and her name was Jalisa. And she was a fashion designer. She was super stylish when I met her. And so she said, and so, but she, you know, she was good at math and science. So she was, you know, a prime candidate

to actually participate in our program. She was already taking advanced placement courses. And so we asked her, would she be interested? And she said, well, I’m interested in fashion. If you can connect it to fashion, maybe I’ll do it. And so we promised that she could, we would be able to connect it to fashion. So she came. And so that summer we were in a repurposed room that used to be an old network server room. So the temperature was super cold. So I don’t know

what happened with rewiring that room, but it was super cold. So every day she would come in with bigger and bigger coats until at some point she came in with a puffy jacket in the middle of Atlanta summer, right? Like it’s just ridiculous. And so she says, I know what problem I want to solve because we gave students an opportunity to solve problems that they were passionate about. And so her problem was solving, helping people who are anemic, who have low blood,

blood iron, you know, who often get cold because of that, keep warm without having to, you know, wear big, you know, huge puffy jackets in the middle of the summer. And so using her fashion design skills, she decided she was going to design a really low profile jacket who had heating elements woven in throughout such that if you just turned a normal button on the jacket, it would increase the temperature of the jacket or decrease it.

And so she created this jacket over the eight weeks of the summer program. She then continued it on into the fall and actually submitted to compete in a regional Grace Hopper conference, which is a celebration of women in computing conference. So she submitted it there as a research project. And she was competing, she was a high school student at the time, actually, a senior high school student, and she was competing against, you know, college undergraduates,

you know, freshmen, seniors and undergrad. And she won first place for her heated jacket, the cozy kid. And so she had the opportunity to then travel to California and go to Grace Hopper, which is the national women in computing conference and present her work there and be honored as a scholar. And she went on to take AP computer science and then to double major in computer science and fashion design. So her story inspires me and really keeps me motivated.

And because there’s so many other stories like Jaleesa’s for students that I’ve worked with that just inspire me to continue with this work and to know that making computing accessible and helping students connected to what they’re interested in is a viable pathway for engaging underrepresented minorities in computer science. Absolutely. And I think that just the opportunity to have that tangible thing that she was able to create, I would imagine just helped pique her

interest. Like it wasn’t something that you couldn’t actually see in real life or reality. This was something that was tangible that she could wear was fashionable, you know, it just really that’s really an awesome story about just the power of what you can do when you’re passionate about something. And you see a problem or a challenge, and you’re using a body of knowledge to help you solve that problem or challenge. So that’s that’s really cool. This project actually

sustained her interest. A lot of times students lose interest in computer science because it gets challenging, but she persevered over the challenges. And I think actually built about three different jacket designs and submitted it to her her her school wide and district wide competitions for the science fair and actually won and was promoted up to regionals. Wow. So it’s a story of persistence. And that’s one of the things that inspires me as well.

That’s really cool. And so speaking of things that kind of inspire you ways that you can think about doing using technology for different facets of work, I’d love to talk about your thoughts on just the future of work and specifically artificial intelligence or AI. And so you are the co chair of the AI for K 12 working group, which is a group that was formed by the Association for the Advancement of Artificial Intelligence and the Computer Science Teachers Association. And that

group is charged with defining for artificial intelligence what students should know and be able to do. And so last December, December 2018, AI for K 12 drafted five big ideas that every student should know. Absolutely. The five big ideas in AI were motivated from a desire to make AI concepts accessible to K 12 teachers and students, because not a day goes by that we don’t read headlines about Alexa or other AI enabled personal assistance. And there’s fear and

concern about drive self driving cars, or the newest AI enabled technology has helped farmers or doctors or scientists solve problems. So we acknowledge that we are in a day and age of AI consumers where everyone has a smartphone or growing number of people are using these personal assistance at home or in their cars. We use Snapchat filters to augment our pictures or Netflix or Pandora to make recommendations about the next new thing we should listen to.

But there’s also this like huge group of DIY do it do it yourself, makers, designers, and citizen scientists who are making use of freely available computing tools, 3D printing and AI services to solve problems that are important to them, and to create products and artifacts that inspire them. There’s also this like huge group of people who are bloggers and YouTubers who are, you know, seeking to share their expertise and opinions. So AI and computing, ethics, etc. aren’t

the privileged domain of like academia or big industry companies or news outlets anymore. These technologies, ideas and concepts are part of our everyday lives and our daily narrative. So more than ever, I really feel like AI is for the people. Oh yeah. It is for the people and so if it’s for the people, how do they learn how to use it? Where does that

start? You know, how do they get access? I mean, for some people, they’re already into technology or they’ve been raised, you know, doing, you know, things in computing. So, you know, you know, kind of jumping into this new domain is something easy. They just watch a video or they read a book or they tinker with it and they get it. But for some people, they actually need a little bit more hand-holding and a little bit more structured environment to learn about AI.

And I really feel like that’s what we’re working on with the AI for K-12 guidelines is we’re saying it’s not for just the privileged, you know, who may find the inspiration to figure this out on their own. This is for people who, you know, within a structured environment can actually thrive and understand and create really amazing things. And so how do we help K-12 teachers and students know how to interact with AI, how to develop AI, and how to understand, like,

where its limits and what it can and cannot do. And so our goal is to create, like, AI fluency, or some people call it AI literacy. And I like to think about it as AI thinking, like, what are the ways of thinking that AI enables for us, right? So computer science and computational thinking allows us to think about problems as something we can break down into smaller parts and something that we can create an algorithm that a computer or a series of steps that a

computer can use to solve a problem. What AI tells us, well, we don’t always have to explicitly program these things anymore. We can teach technology through examples, you know, AI through examples, how to create their own solutions to these problems. And leveraging that as a tool to create, you know, solutions to a wide range of problems that we haven’t even begun to even think about. So let me unpack some of these big ideas for you. But I just want to give you a little

bit of context about where we were starting from and what really was motivating us to get into this space. So the first big idea is about computers perceive the world using sensors. So it’s about perception. It’s about computer perception. Like, we have human perception, right? We have our senses. We have our eyes, our nose, our ears, our tongues, right? Or we can feel. And so these human perceptions allow us to take in information about the world and to understand and interpret, you

know, what’s going on around us and to have interactions with people. Well, computers use sensors as well. And that’s what this big idea is about is how do they use sensors and how do they make meaning from those sensors, right? So just because they see a picture of a cat or maybe they see a stop sign doesn’t actually mean that they know what it means. How do they make meaning of what’s in that picture? And so perception is about the algorithms and the process by which

computers make sense out of the things that the data that they pull into their sensors. It also bridges us into this idea that, you know, computers and AI, you know, they perceive differently than humans, right? So we have our perception system, our senses are connected to our brain and our brain is what’s doing the reasoning and the computation around making meaning. Our prior experiences help us make meaning. So how do computers do this, right? So computers, they don’t

have, their brain is the computer, right? The CPU, right? Sensor processing. But essentially programmers are designing the framework to help the computer to perceive and understand, right? They’re helping the computer know, oh, that, you know, octagon you see is actually, that’s red, is a stop sign, right? That roundish thing that you see with, you know, two specs close to the top, oh, wait, those are eyes and that’s a forehead and that’s a mouth, right? So it’s helping them

understand, you know, what is it that they’re seeing? Not just identifying, you know, blobs in a picture. It’s actually helping them identify. So I think one of the core ideas of these big, the big ideas in AI is that there’s a role for humans to play in this, in AI and there’s a role for the computer to play. And that there’s this symbiosis, symbiotic relationship that’s happening. It’s not just AI is rogue and doing whatever it wants. AI, you know, for the most part,

actually operates in a pretty confined space on solving pretty narrow problems that, you know, AI designers or developers or people who are employing AI use to solve problems. So perception is about that. And so we’re thinking about, well, what is, what does this big idea look like for K2? Well, for K2, you know, maybe students are just learning the difference between the sensor at the supermarket and, you know, the camera sensor on your laptop that opens,

or that opens when it sees your face, right? So that’s really what the first big idea about perception is about. The second big idea is about representations and reasoning. So helping people understand, well, what is it that the computer has that helps it make sense and helps it, you know, make meaning, right? And so computers have representations kind of like we have representations, right? We have like ideas of mental models of what a house looks like. If I asked you to describe

a house, you’d be able to tell me the shapes that go into a house, right? Or if I asked you, you know, what your last meal was, you’d be able to describe it. So these are representations that you have in your own mind about the things you’ve interacted with in the world. Well, computers need to have representations like this as well. And so it looks like a little bit more like what we call in computer science data structured. So ways of storing information in structured ways. So it might be

trees or graphs or different ways of storing information and reasoning is the way that it navigates and accesses the information in these representations to make meaning. And so a common way of thinking about representation and reasoning is to think about how does a computer chess, you know, know how to play against you and know what moves to play, right? It’s not because it’s programmed in every time this person makes this move, make that move, right? It’s not so hard

coded like that. It’s actually very dynamic where it’s thinking about if this player makes this move, what are the possible moves that I can make and what are the moves that I can make after that and after that that’ll lead me to a winning game state. And then choosing from that collection of moves to say, oh, I’m going to pick, you know, this one because it gives me the most possibilities of winning, you know, two, three or four turns down the line. The third one is about learning. So

computers can learn from data. And so we hear about machine learning and deep learning and neural nets all the time in the news and on TV, right? That mean and that’s what this big idea is about is demystifying what is machine learning. I think there is a conflation between machine learning and AI. So artificial intelligence is the broad field and machine learning is a set of techniques that are used within AI, right? So it’s a sub domain of AI. And so machine learning is really kind

of awesome, right? If we think about how we learn, right, like we learn by seeing examples of things as we’re driving down the street and you’re in the car with your son, right? You may see a car and at some point he’s going to be like, oh, car and then he might see a truck and he might call it a car and he might see a bus and he might call it a car and you have to teach him that, oh, no, you know, this is a car, but that bigger thing, that’s a truck and this really long thing

here, that’s a bus, right? And teach him about these different things. And essentially that’s what we’re doing in machine learning is we’re classifying data and giving it labels. And then we’re using those labels and that labeled data set to actually train a machine learning algorithm to know the difference between a bus and a truck and a bus, a bus and a car, right? And so this is a really important big idea, one, because it’s getting so much airplay, but also because I think

there’s a lot of mystery around what’s really happening here. And so our goal with this is to help students understand, you know, what role do AI designers and developers play in creating machine learning algorithms or systems that use machine learning algorithms and who’s responsible for the decisions that they make, right? There’s all these conversations about bias and, you know, bias in AI algorithms, where does it come from? Does it come from the algorithm? Does it

come from the designer? You know, where does it come from? And essentially, you know, there’s this principle of garbage in, garbage out, right? And if you don’t have a good data set, a computer can’t make good decisions. So if that data set is not labeled really well, or doesn’t take into the consideration all the types of questions that someone might ask that data set, it’s not going to be able to answer it very well, right? If I have a data set all about cats and then I start showing it

dogs, it’s not going to have any clue what I’m showing it and it’s not going to recognize the dog very well. And it’s not any fault of the dog. It’s not any fault of the algorithm. It’s the fault of, you know, the fact that we don’t have enough data to make decisions about the things that we’re now being presented with. And so getting students to actually be able to build, you know, or modify some AI system that uses a machine learning algorithm, maybe through training it, so training the

classifier so it learns about different examples, or actually, you know, trying it out in the world and seeing its limitations. I think people will come to understand the interstices and the complexities of machine learning and be able to appreciate that it’s not this omniscient thing. Understand the sources of bias and not, you know, feel like they need to blame people for not being inclusive. You know, sometimes it may be a case of someone’s not being inclusive, but a lot of times it’s

what’s more convenient. These data sets are, you know, really large. You need really, really large data sets to actually solve problems and reason about them. And so if you don’t have a large data set of maybe, you know, African-American faces or hands or things like that, you may not be, or women’s faces or hands, right? You may not be able to accurately identify them. And so that just tells us we need to be building better data sets. Yeah, and it speaks to the opportunity that you were

talking about previously, just about the role that human beings have to play in all of this, and really stresses that even more, right? So, and I think that, I think that’s too, where people get confused to your point. It’s, you know, not this omniscient, just actor, you know, operating rogue in the world. So there’s really opportunity for human beings to play a significant role in all of this. And I think it’s really incredible that these guidelines that are coming out, because we have to

start young with students understanding that. Absolutely, absolutely. The last two big ideas are really big. The fourth big idea is about natural interaction. So, you know, making agents naturally interact with humans and understanding that this is a challenge for AI designers is non-trivial. We have things like natural language processing, which is, you know, trying to parse our texts, like if we were to use the audio from this call, you know, parse our texts to understand

what are we talking about, and what ideas are connected to which ideas, and helping build context, right? Because we as humans, we’re really, we’re actually really powerful computational agents, right? We can have this conversation, and then, you know, you, at the start of this recording, you know, pulled back all the way to a conversation we had two weeks ago, and you were able to remember that. Well, when I interact with Alexa, it’s not doing that. Alexa can barely remember what I said

two weeks ago, right? Because that’s not how she’s programmed. That’s not how she’s designed. She’s designed to be a transactional agent, whereas humans are conversational agents, right? We’re maintaining context. We are, we are able to understand a lot of ambiguity in conversation based on our context, and based on our prior experience in history with that person, or that topic, to begin to, to hold engaging conversations. And so this fourth big idea is about how do we do that?

You know, how do we get computers to be able to hold these type of fluid conversations where it’s not transactional, like Alexa, or Google, or whatever, where, or Siri, like where we have to give a wake word, and then a request, and if you don’t have that request formulated, well, it will not carry out that request. It will like, I’m not really sure what you’re asking me to do. Try again. We don’t, if we had those types of conversations with our family members, we’d be infuriated.

So how do we make computers have these conversations? This big idea is also about things like emotional intelligence, right? Like, as humans, we have a tremendous set of resources to interpret when people are sad, or angry, or happy, or excited, you know, we’re listening to their voice, we’re looking at their face, we’re reading their body language. Well, how do we teach computers to be able to do that? And to be able to do that well, right? Because if you’ve ever been in a

classroom and you’re teaching, you could sometimes be fearful that you’re boring students to death because their face are just blank stares. But then, you know, when you pause to ask a question, that same student that was looking with a blank stare asked this amazing question or makes this great comment and realized their face was just, you know, looked blank because they were processing and thinking and concentrating, right? So, you know, there are all these miscues that someone can

read, you know, about people as well. And so how do we teach computers to understand those things? You know, there’s things like sentiment analysis that would fall into this category, which, you know, is being used right now to detect bullying or depression and things like that in social media posts. And so that’s, we’re actually doing fairly well in that. The emotion detection, we’re growing in that area. But to do it fluidly like a human, we’re not quite there yet.

And so this big idea, like I said, has a lot of stuff in it. There’s stuff about collaboration, like how do you collaborate with robots or AI agents? How do you work cooperatively with them? What does that mean? You know, how do you, you know, how do you design a robot such that, you know, it stays close enough for you to ask it questions and for it to respond and handle things you need, but not so close that if you decided to stop abruptly that it would be running over you

and the two of you would be on the ground, right? So it can give you that social space that we as people are cognizant of and understand. This idea also gets into this, you know, something that’s fairly controversial, which is about consciousness. You know, can AI have consciousness? So it’s kind of philosophical. It’s been a longstanding debate for many years. This is like the crux of the test from Alan Turing, which is actually, you know, about can a person, you know, detect if it’s a

computer or a human that it’s talking to? And what does that mean to be conscious? And what does it mean to be intelligent? So the idea that, you know, AI can one day be sentient in the way that humans are sentient and aware is this concept of artificial general intelligence, where it has the same level of intelligence as a human possibly more. And that we’re really far off from that. Like that is, that’s, I mean, yeah, we’re really far off from that. If you talk to any AI expert, what they’ll

tell you is that what we’ve done really well is narrow artificial intelligence, you know, domain specific artificial intelligence. So we’re able to detect retinal neuropathy, which is a disease of the eye caused by diabetes. We’re able to detect that really quickly from a cell phone and really accurately, much, much faster than a human, because there’s specific markers that can be detected. And we can do that well. We can play chess, we can play go, we can play these things

that are a little bit more structured and confined. We can identify cats or dogs and pictures or faces, because there’s a well defined structure to them. But those are very small domain. To be able to have general intelligence, where we’re combining all the levels of intelligence that we have from our sensing and perceptions of the environment, our sensing and perception of other people and interpretation and making meaning of what they’re saying and putting that all together

to become what we are as a human and putting that into an artificial robot or agent, we are not anywhere near close to that. But the individual things we are growing rapidly and very fast on. So this big idea is really about helping students understand one, the difference between human and artificial intelligence, and two, what’s working well in artificial intelligence and what’s possible right now. Yeah, there’s a lot that’s coming out too. And just thinking about the future of work

that is talking about what you’re speaking on right now is just those uniquely human characteristics that AI may someday well in the future have. But just how are you, how are we thinking about preparing jobs for people to continue to do those uniquely human elements such as compassion and these sorts of things that in the question two would be like, do we want AI to kind of venture into those sectors that are based upon uniquely human characteristics? So I think that’s a really

important understanding for people to have to is that we’ll still need people to feel to be able to have those things that make us in our very nature human, right? Absolutely, absolutely. And that’s what the fifth big idea is about. It’s about social impacts. It’s really being able to have discussions about what are the beneficial or harmful effects of AI in society, right? Like build something that you think is going to help. But in doing that, it also harms another

part of the economy. How do we address that? So that’s really what this fifth big idea is about. It’s about thinking about the ways that AI can be applied to solve problems. Like I told you about the retina neuropathy diagnosis in third world countries. It could be about self-driving cars. It could be about how to use AI systems to detect bullying or to make sense about prison sentencing. So we get again into the ethics and bias questions here. We get into, you know, if you

can make it, should you make it? We get into those conversations about, you know, if you can make it, you know, how do you make it such that it’s accessible for all people? And so you begin looking at the applications, the implications, and, you know, what are we going to do to mitigate any harmful risks that may arise because of that? And so I think this fifth big idea is going to give us lots of fruitful conversations and hopefully prepare students to have these kind of conversations

with their friends and family outside of the classroom, such that we can decrease the fear that people have around AI and begin to think about it a little bit more soberally and in perspective about what is possible and then, you know, what is it that we should be doing or can be doing with AI? Such that, you know, at some point, you know, the students that are in, you know, class of 2021, 2022, right, they’re going to, or even 2050, are going to be voting on, you know,

laws regarding AI. They’re going to be voting on, they’re going to be in juries deciding, you know, if an AI agent is guilty or if a human agent is guilty and who should be taking responsibility and they need to have a basic knowledge about how AI works and what’s possible and who’s doing the designing, who, you know, what things are black box, what things are intentional to be able to make these decisions wisely. Yeah. And that kind of leads to, I mean, the one question that comes up

in my mind then with these big ideas and trying to get students and their teachers to understand more about AI, more understand more about these fluency and ideas that you guys have laid out. The question is just for me, how, like the paper ends with the call to the AI community to provide opportunities for students and teachers to learn more about AI and how it will shape their future, but how do you envision that actually happening? How do you see the community, like getting into

schools and teaching about AI, even if it’s not organizations or businesses who are helping shape this, how do you think that it’s going to shape education and actually translate down into a teaching and learning opportunity? Well, I think that there’s been a lot of work in the CS education community around getting computer science into the K-12 classrooms. Our goal isn’t to necessarily create a whole new curriculum and a whole new set of guidelines that, you know, now have to be

implemented as a standalone course, but is to piggyback on the efforts that are already going on in the CS education community to get computing into K-12 classrooms and to train teachers. So with the guidelines, we are not creating curriculum at all. What we’re doing is we are creating guidelines or what some might call standards in other communities to help teachers and students know what is important to know about AI. And we can because we’re working with 16 K-12 teachers

for each grade band is helping to figure out what are the connections back to things that they would already be learning in their math, science, and social studies class or English class, and what might they already be learning about computing and how these can be integrated in because we know teachers don’t have a tremendous amount of time to build new lessons or to cover everything, right? Because there’s so much in the curriculum to cover it already. So looking at ways

that we can actually piggyback on what’s going on with existing efforts. So that’s our first way. Our second way is through actually creating these guidelines. So creating guidelines that actually explain AI, what should students know about it? So understand core concepts and foundational concepts and each of those five big ideas and also what should they be able to do? So that creating that because I believe if students create things and learn skills, it empowers them to use these

skills and tools in the future outside of the classrooms and in the community. I think for me a big part of the guidelines are about what students should be able to do in empowering students and giving them the tools. So that brings us to our third objective, which is to build a community of AI resource designers and developers. So people who are designing curriculum, designing activities, designing tools that bring AI to life for students that allow them to look under the

proverbial hood of AI and understand what’s happening in that black box. What’s really happening so that way it’s not magic. It’s something that they can understand, that’s understandable that they can reason about. And so that’s our goal is to build this resource library, also build a community that can help us fill this resource library. There are a tremendous amount of tools that are out there already that are amazing. They demonstrate opportunity, they

give students opportunities to build classifiers, to classify data in trained machine learning algorithms, to recognize a whole sort of things from pictures to images to sounds to anything you can think of. So they allow you to do that, but right now they don’t allow you to look under the black box. So we’re making a call out to the AI research community that says, hey, help people look under the box of how these neural nets work or help them look under the box about what representations are in a Netflix

recommender or how does the Snapchat filter work? Help them understand what’s happening, what is all the processing that reasoning the computation that’s happening underneath these things. You can do that videos, you can do that through interactive activities, you can do that through unplugged activities, so things that don’t even require the computer. And providing a host of ways for teachers and students to access these ideas. And I just want to point out too that you

mentioned all of the resources that are being collected as a portion of this and that is all available on the website which is ai4k12.org, right? The resource directory. And so that is really great. And you guys are doing a host of events throughout the rest of this year where you’re talking more about this. And then ultimately, this will all result in those national guidelines, which when do you think those will come out? We’re hoping to have a draft of them this summer.

And I say draft because with any new project that’s starting from scratch, you know you’re going to have to go back and iterate. So our goal is to have a draft out by this summer, but we know we’re going to be doing some more iteration. And as teachers are becoming more comfortable with the stuff and trying it out, we’re probably going to do another pass on it fairly shortly after just to kind of clean up some stuff loose ends that we may have missed. And

as we’re engaging, you know, both the teacher community, the CS education community, and the AI research community, being able to understand if we have gaps and getting that feedback along the way, it’s one of those things we’ll be doing as well. I wanted to just get your thoughts to Christina because you mentioned a little bit of this when you were talking about Jalisa’s story. But we know that unfortunately that there are some girls and students of color who

may be intimidated by computer science either because they don’t see themselves represented in those fields or because they think that they’re not good at that type of work. And so how are you working to change those perceptions? And what would your advice be for students who are feeling that way in particular? I think for me, one of the ways that I personally try to break down those stereotypes and perceptions is just by being visible. So going to speak to

students in their classrooms, hosting these summer camps and after school programs, you know, making sure that the students that are in my lab are representative of the types of communities that we want to go into. And that people are aware of the issues. I think a lot of times, even though we talk about it, there’s a lot of people that are so very naive about the barriers of entry for women and underrepresented minorities and computing. And so, you know,

partially doing a little bit of advocacy work where it’s like where that opportunity comes up to say, hey, you know, we should be doing this, we should be making sure that we’re getting into minority or under-resourced communities and schools. So I think that’s one of the ways. I think for me, what’s most impactful isn’t just going and telling students they can do it, it’s giving them the opportunity to do it. And so that’s why for me, these after school and summer

camp programs are really pivotal. I think if it wasn’t for me actually being able to create something, if it wasn’t actually for Jalisa to be able to create something and connect it back to our passion, she probably never would have pursued computer science. So I think that there is a really big need of being able to help students explore the field, not just as an outsider looking in, but from an insider, you know, pushing out. One of the students that I worked with when I was at

Georgia Tech, and I tell these stories because they’re at the point now where they’re mature and they have jobs in computing, right? And I think, you know, seven years ago when I started working with these students, I wasn’t sure where their story was going to go. I just knew that I felt like I was doing the right thing and I felt like I was making an impact. But it’s now that I look back on their lives and their trajectories and go, yes, it was that program was one of the pivotal

points in their trajectory that I feel like these stories need to be told. So when I was at Georgia Tech at the same time, I was working with Jalisa actually, it was the semester after. It was that fall. We decided to move it from just a summer camp to a semester long after school program, which actually ended up being a year long after school program. And we met Amanda. And Amanda was, you know, from, you know, a rough background. And she was really street smart and savvy. And she

just wanted a job that she could work in education that will allow her to increase the opportunities for her family and herself. And so she hadn’t been involved with any technical after school programs or clubs or anything. But somehow we got her to participate in this program at Georgia Tech. And she came to us and she, we always, like I said, we always have students think about what is it that they, what problem would they like to solve? And for her, she wanted to solve problems for

families with deaf children. So hearing parents, but that had deaf children and that we’re needing to learn sign language and understand how to better communicate with their children. And so I said deaf, I meant blind. And so one of the things that she wanted to do was create a braille, a tablet based braille reader that would allow, you know, parents to read stories to their kids, but also provide the braille for the children and also help the parents actually read braille.

And so, you know, as part of any project, you got to break it down into parts. So the first part she was working on was doing a braille, was doing a text to braille translation. And so one of the ways she wanted to represent the braille was through LED lights. So, you know, little small lights that she’d arrange in patterns that when they lit up, they would actually, you know, create the braille pattern and then show beneath that the text that went along with it. So that a parent and a child

could read a story at the same time. And so she created that and, you know, she struggled a little bit with the programming. She wasn’t as passionate about the programming, but she felt like, man, this electrical engineering stuff, building these circuits, I really like this. This part, doing things with my hands, this is what I can do. And so she got excited about that part of the project. And she went on to present for AT&T at the time, was running a competition using the Arduino board

for students to create devices and actually, you know, pitch those products to AT&T executives at the time. And so she, I believe her team won second place in that competition. But that set her on a path to become more interested. So when she was a junior at the time, so she started taking some computer science classes the next year. And then she kind of struggled through those a little bit. But then she was like, but I still like this stuff. So she decided she was going to go on and major in

computer science. Well, fast forward, she has majored in computer science, graduated with a degree in computer science. She has interned at companies. She’s done research at several different national labs and universities. And now she’s a PhD student at University of Alabama. And it’s just amazing. And she’s a PhD student in computer science. Yeah, that’s awesome. And so I think, you know, to double back on your initial question about perceptions, I think there are things students

believe about themselves initially, one, because, you know, they don’t necessarily, they’ve never explored the area. So they may go, I don’t really have an interest in it. But we give them an opportunity to explore it. They go, hmm, I’m interested in this part, but not that part. If you keep them engaged, and you sustain that engagement, it allows them to explore, you know, the depth of their skills, actually build skills. And to, you know, in Amanda’s case, she, you know,

she ended up building up skills in an area that she was weak. And she didn’t really understand somewhat a little fearful in the programming to mastering that and that being part of her major, part of her career trajectory. Right. And so what I guess the message to me, and I hope to everybody else is that don’t give up on students. That’s because they don’t look like they’re quote unquote getting it, or they may not be into it, you know, right away, give them some time, give them some

space, sometimes the initial perception and persona they give off of, Oh, I’m not interested in this is because they’re fearful of it, and they’re not sure how well they’re going to do. Give them that space to fail softly and quietly and to get feedback, you know, it builds their confidence to the point where they’re like, you know what, I can do this, and I want to do it. And that’s that they’re making the choice, not us. Yeah. That’s such good advice. And I think just from

the standpoint of where it all starts, it starts with allowing them to explore and base it on a project based learning standpoint or problem based learning, just really having inquiry about this topic area, which is what we recommend in education generally, right. And so therefore, it should apply to when you’re talking about computer science as well, if you have the opportunity to actually do something. And if it’s based upon something that you’re interested in or passionate

about, it’s more likely to stick and keep you engaged. What are your predictions on the future of work and how it’s going to impact people and communities? Well, there’s several things. You can have this conversation probably a hundred times and come up with new things. Right. But I think one of the big things that that occurs to me is that we don’t really know what the future of work is going to be, right. Because when I was in

middle school, there was, we weren’t thinking about the internet. Yeah, you know, it was really about the internet. We definitely weren’t thinking about social media and that I would have Facebook or Twitter or any of that stuff. And so I think a lot of the future of work and a lot of the future of the technologies that we’re going to be interacting with are the seeds in these children’s bellies right now. And we need to create fertile soil and water and cultivate

the dreams and visions that they have. Because that’s the future of work. That’s the future of our society. And I know that sounds like super philosophical, but I fundamentally believe that I don’t even think I have enough imagination to imagine what the future of work would be in that context. But what I can say is that what’s going to be really important is that students be agile, right, that they understand that yes, I’m going to get this, you know,

you know, 13 years of experience in these classes in K-12, I may even get an undergraduate degree or some, you know, or some graduate level degree. But through all of my education, there’s going to be stuff that’s going to be taught to me, there’s going to be stuff that I’m going to have to be reaching out and looking at what’s changing around the world and what is it that I’m curious about and need to learn about. So I’m always a lifetime learner. I think if we cultivate any skill in

our student and we’re going to, if we had to, you know, choose something, that would, that’s what I would choose. I would choose to invest in helping students understand how to learn and how to continue learning and how to make decisions on their own, right. I think a lot of times we try to teach students how to solve problems in certain ways and we’re starting to break out of that now, you know, when we’re thinking about inquiry learning and project-based learning and problem-based

learning, we’re trying to teach students how to ask good questions and give them the tools to answer them. That’s the key. Any final thoughts that you have, Christina? I think for, you know, just a message to all the students and teachers, you know, is that when you’re, when you’re teaching these students, it’s helped them realize that, you know, they may not see the job and the career they want right in front of them right now, but to imagine what that, that job or that career might be in

the seek after it. That is wonderful advice, Christina. Thank you so much and thank you for joining us today on the Getting Smart podcast. Thank you for having me and allowing me to chat so long. A big thanks to Christina Gardner-McEwn for speaking with us today. We love how much she stressed that one key to getting students hooked on computer science is to really allow them to follow their passions and drive their learning based on something they’re interested in or propose a

problem or challenge that they can solve with technology. And we couldn’t agree more with her that the future of work will require all of us to be lifelong learners. For more thinking on the future of work, be sure to head to our website and look for the future of work series. You can also visit gettingsmart.com slash future of work. Be sure to check out the Getting Smart podcast on iTunes as well. And while you’re there, hit subscribe and leave us a rating. And for more on

all things innovations and learning, check out our blog at gettingsmart.com. That’s it for today listeners. Thanks for tuning in. For the Getting Smart podcast, this is Jessica signing off.