"Team Member, Not Team Leader - Teaching Kids To Be Smarter Than AI"

Every day in classrooms around the world, teachers are watching students reach for AI before they have even attempted to think for themselves. Not out of laziness. Not out of malice. But because having the right answer at your fingertips in five seconds is a temptation that even adults struggle to resist.

In this episode Florence sits down with Sara Refai, Head of the High School English Department at Saigon South International School and a teacher with over 15 years of experience building safe, collaborative spaces where students learn to think for themselves. Sarah is also the kind of teacher every parent wishes their child had.

Together they explore one of the most urgent questions facing parents and educators right now — how do we protect our children's ability to think critically, sit with uncertainty, and find their own answers in a world where AI will always offer an easier way out?

Sarah shares the classroom exercises she uses to help students see AI as a team member rather than a team leader. She reveals why teenagers' strong sense of justice is one of the most powerful tools we have in navigating AI fairly. And she is honest about the moments that frighten her most — including the growing number of teenagers forming emotional relationships with AI chatbots outside the classroom.

You will also hear why failure matters more than ever, why the temptation of AI is not a moral failing but a design feature, and what parents can do right now to push for better AI education in their children's schools.

This conversation will change how you talk to your child's teacher — and how you talk to your child — about AI.

🎙️ Listen on Apple Podcasts — podcasts.apple.com/gb/podcast/the-ai-mum-podcast/id1892893748

Follow on @theaimumpodcast on Instagram

Transcript

SPEAKER_01 0:05

We stop teaching our children that it's okay to fail, it's okay not to have the right answer, it's okay if it doesn't look as good. You see what I mean?

SPEAKER_00 0:13

100% when you have the correct answer at your fingertips, it's easy to underestimate that failure is important, and especially that it doesn't feel good. It's totally the opposite of what AI is doing. You almost have to give them the cue to say, no, your thoughts matter. The best case scenario for AI is that they learn to use AI as a team member, not a team leader.

SPEAKER_01 0:43

Welcome to the AI Mom podcast. Today we are going to focus on how teachers can forge critical thinking for their students when it comes to AI and how important their role is in making sure our kids see AI for what it really is. With its positive but also its downside. After all, our teams are going to be the ones shaping the future, and the question is how can they do so when their reasoning and critical thinking skills could be threatened by the rise of AI in their lives. Let's get into it with our guest, Sarah Rifai, head of the high school English department at Saigon South International School. She has over 15 years of experience in building safe and collaborative spaces and supports her students in developing their critical thinking skills across various topics. Most importantly, she's a dear friend and the kind of teacher you really want for your child. Hi Sarah, thanks for being with us today. Thanks for having me. So the first question I wanted to ask you is give us an overview of how AI is present in your classroom.

SPEAKER_00 1:56

Yeah, to be honest, it's become so ubiquitous. It's almost become a question of when is it not present? And how do we even put in guardrails? For a lot of students, they don't even Google things, they just AI it. So as soon as there's a laptop present or a piece of technology present, the reality is, as the teacher, you can't control when they are thinking and when AI is doing their thinking for them. So we have to become much more structured in thinking about how we manage our classroom space. So if a discussion question comes up, some students, their instinct is to ask AI, but for me, I want them to come up with their own ideas. So I have to make it a point to say, like, no, don't access AI. I want you to think about it first. And you almost have to give them the cue to say, no, your thoughts matter.

SPEAKER_01 2:48

And do you think they're confident enough to give their own idea without AI?

SPEAKER_00 2:53

I think that you have at your fingertips an easy answer that's gonna sound good. So if we're a student, it's A, maybe you don't have the confidence, but B the temptation of having your work done for you is really powerful. Like I'll give you an example. I might pass an assignment through AI just asking for general feedback. And AI will respond to me and say, hey, it looks like you're writing a paper. Do you want me to write it for you? Okay, I'm 46 years old. Even I was like, oh my gosh, like in five seconds, I could have a 15-page paper. So a 16-year-old, a 12-year-old, an eight-year-old student who's stressed out, they're scared their teachers will be upset at them, they're scared their parents will be upset at them. The temptation to have this work done for you doesn't come from malice. It's so easy, it's so tempting, even for adults.

SPEAKER_01 3:56

I think you're touching on two points here. Teenagers and children are not set up to stop this temptation. It really means we have to do something about it. And the other point which is interesting is the fear. Because given AI always has an answer, we stop teaching our children that it's okay to fail, it's okay not to have the right answer, it's okay if it doesn't look as good. You see what I mean?

SPEAKER_00 4:20

100%. Learning comes through failure, learning comes through challenge, learning comes through mistake. And so when you have the correct answer at your fingertips, it's easy to underestimate that failure is important, and especially that it doesn't feel good. I really focus on the idea of if you can come to the table with something nobody else has thought of, then you're adding value. And that comes from noticing, observation to detail, trying, sitting in the discomfort of not knowing.

SPEAKER_01 4:51

It's totally the opposite of what AI is doing. And what's their response to that? Do they prefer the certainty of AI? What do you see from them?

SPEAKER_00 5:00

A lot of it, I think, comes from our habits of thinking that we teach from a young age. If they're used to this, then they're much more comfortable and familiar with the absence of knowing. And they have confidence in knowing that they can reach the other side. They can get to a place where, if they've thought about the evidence, the information, the data for long enough and hard enough, they'll see a pattern, they'll find an answer, they'll find a way forward. If they've been brought up with a lot of rigid, very binary thinking, then there's a real fear to not knowing the answer. And I think that vacuum is where they need to get comfortable.

SPEAKER_01 5:40

So, how do you make them enjoy the journey of finding something by yourself? Because that is also rewarding.

SPEAKER_00 5:48

Yes, and that's such a good question. And in a way, that's always been, but the core of education. Learning isn't always fun. How do we make it engaging, compelling to the many students that we work with? One of the key things I think is allowing a certain amount of student choice. We have a structure. We're trying to get to this place. How we get there, there's a lot of flexibility. So, for example, my 10th graders, they do a digital media literacy unit. And at the end of it, I want them to use modern tools to explore how different social media platforms are being used to disseminate news stories and the positives and negatives of this misinformation, disinformation, bias, etc., etc. So that's what we want to get to. But how we get there can really be their choice. What tool do you want to use? Do you want to use TikTok, Instagram, Reels, YouTube, shorts, infographics, articles, photojournalism? And then what do you want to talk about? What's something that's interesting to you? So allowing them as much as possible to pick the things that they're anyway curious about and hooking them through that curiosity into a skills-based progression is kind of how we try to do it. It doesn't always work, but I think it works more often than it doesn't.

SPEAKER_01 7:16

I love that actually, because here you're combining curiosity and freedom. You mentioned this exercise with the media. Do you have something specific you go through with them to discuss AI, to discuss the answers, the accuracy, the human-like feature, what they think about it?

SPEAKER_00 7:34

So two things that I like to do with them. One is I like them to really start to learn that the best case scenario for AI is that they learn to use AI as a team member, not a team leader. And that they shouldn't outsource their thinking to AI, but AI can be leveraged to help them. So, okay, now AI is your team member. That means you might give it some work to do, but you're going to check that work and you're going to verify and validate it. One exercise I enjoy doing with them is having my students analyze a piece of text, let's say, for example, an advertisement for a beauty product. How did the creators of the advertisement use persuasive tools to convince someone to buy this product? Maybe they tapped into certain beauty standards, they used certain colors, they used a model that looked a certain way, or a celebrity andor, etc. Once students have done that work and they've discussed it and they've really deconstructed it, then I asked them to put that same image into AI and get AI to analyze it. Then a comparison. What did you notice that AI didn't? Why do you think that happened? What did AI notice that you didn't? Why do you think that happened? And that gets them to start evaluating it as a tool. And oftentimes they'll notice certain human elements, certain subjective elements that AI just can't navigate or read. But they'll also notice that AI is very good at certain things, like picking up very small details of text. So having them already starting to evaluate and see themselves as good as AI. I noticed things AI didn't notice. It can't do what I do. That's so empowering. Empowers them 100% and reminds them that AI can't do it for them, it can help them.

SPEAKER_01 9:32

The concept of team member, not team leader, I think is such a great thing to be telling them, even to be telling ourselves. And I like your approach, which is nuance. There are good things, take them, but bad things, notice them, and you are better than AI still, you know, because I think there's this competition right now between AI and humans, and even teenagers must feel a bit vulnerable about it.

SPEAKER_00 10:00

And I talk to them about how I use it. For example, I'll say I used it to create a quiz, for example. But I'll say, you know, it made a mistake. So really having them use it and pay attention to when it's holding them back without them realizing, and when it's opening up avenues for them.

SPEAKER_01 10:20

In our previous discussions, you mentioned something that really stayed with me is that teens have a sense of justice that is really strong. And I think that's a great thing to build on because we have to make them realize how important that is and that they are right in having this strong sense. So if you could just take us through this again, because I think it's so important.

SPEAKER_00 10:44

One of the things that has been increasingly difficult with AI is when we think of the practice of assessment or testing, and how do we ensure that academics or grades or whatever system we're using are rewarding students fairly. And students feel very strongly a sense of injustice if they think that someone is being rewarded unfairly and they've done the work and they've tried, and then somebody gets a higher grade or someone gets a university acceptance or something, and they feel like that person didn't deserve it. So navigating the difficulty of how AI is changing our classrooms and changing assessment in particular, and trying to make sure that there's equity for our students, both in the classroom, and if you think globally, even like the digital divide, right? Who has access even to these tools and who doesn't is becoming trickier? So you see, for example, some schools or some teachers will go back to paper and pencil tests and timed. And there is a space for that. But putting all of our eggs in one basket or only doing one type of thinking, only doing one type of assessment will hinder someone down the line at some point. So, how do we have a diverse range of assessment that is fair, that is clear, that is transparent, but allows there to be this kind of equity and ensure that there's as little injustice as possible just in terms of how people are accessing these tools. And it is a very tricky conversation that's still ongoing in the field of education for sure.

SPEAKER_01 12:33

Even in the industry, because right now there are different types of subscription you can get with these AI chatbots, which are more efficient as you pay. And I think that's also not fair for teenagers who don't want to pay or cannot pay or who don't have a parent who has a subscription, for example. I think it's just good to have classes that are without AI and classes that are with, so that teenagers can shine in both worlds.

SPEAKER_00 13:01

Yeah, I think that's a really good point. And I think in a way, because it's such an ever-changing landscape and it moves so quickly, having that flexible adaptive approach will allow us to keep up a little bit better as well.

SPEAKER_01 13:19

One thing I'm very interested and concerned about AI and children is the human-like features in AI chatbots. Did you have situations where the child came to you and said, Okay, I need to talk to a real person? Or do you think they're hiding more behind their AI tools to be able to speak about vulnerable things? Do you have any insight in that?

SPEAKER_00 13:43

And I think the point you touch on, which is so concerning and so frightening, is that there's so much we don't know about what happens outside of our classrooms, outside of our homes, like AI is on the phones. So I haven't yet personally dealt with this, but there are studies that show that a huge number of teenagers either are or have been in a relationship with AI. So one thing that we can do as adults is teach students what it is that creates the AI tone and that relationship-building sense. So, for example, one thing I do to keep it kind of academic is I teach students how to use AI to get feedback on first draft writing, which is a very kind of basic low-stakes thing. But we talk a lot about tone. And we talk about the feedback, and I ask them to work with AI in different ways. So, first, because I want them also to think of it as a team member, how to ask it to change its tone to a tone that works for you. Not everyone enjoys the AI tone as it is. How to talk to it about your cultural background so that it can have culturally appropriate communication with you as well and norms. But an exercise that I do with them at the end is I ask them to do this across different platforms and see what feedback they get, what grade they get. And they'll notice it's never the same. But then I challenge them to say the first person to get AI to change their grade wins, for example. Within 30 seconds, I'll have six or seven hands gone up who have managed to get AI to change their grade. And then we talk about the why. Why did it change the grade? And then we get into the people pleasing. We get into it's not treating you as a person, it's treating you as a customer, you're a client. It's trying to create engagement. And so it starts to fit into how we talk about an algorithm, because we talk about the algorithm in social media, how it's trained to keep you engaged and keep you on. And so I think when we talk to them about what creates the tone, why does AI sound human? And we really actively create activities that show them the motive behind the tone. I think those are some of the ways that we can empower them.

SPEAKER_01 16:08

I think every school should have an hour a week discussing these topics with the children. I even wonder at what age you should start having these discussions because right now you can see that it's not only something that is for teens, children from nine years old, they're using it already. Even in their Google search, but they're finding an AI answer. So for me, it's so critical to have first teachers like you, but also classes like yours early on. I think it's so empowering, and it should be in the curriculum of all schools. And also, parents can do something about it by pushing for it, asking for it in the school of their children. But before we leave, I have one last question. What is the biggest downside of AI, but also what is the biggest upside of AI for our children, in your view?

SPEAKER_00 17:00

Yes, for sure. And I'm so glad that you have these two elements to this question. So the biggest upside for me is that AI can really help students with things like self-management, scheduling. If they have a big project, for example, that they have to do, they can use AI to help them break it down, to help them schedule their tasks, to help them get low-level feedback and to guide them across the process. So it can be very helpful in that way. I think the downside is when it goes too far and they're no longer thinking, the dangers of the validation of AI creating an emotional dependence on this thing that will talk to you for hours and validate everything that you say and allow you to go in through the minutiae of your life. So, yeah, I think that's it. I think that it can be very effective. And then it's as soon as it goes too far, it becomes the opposite.

SPEAKER_01 18:00

Yeah, 100%. Ultimately, it's a tool. So, like any tool, it can go wrong, but it can be also very powerful. So I really hope that with time, either we have AI chatbots that are more customized for our kids, less human-like, less following up by a question every time a question is being asked to them. I'm hopeful the children themselves will create their own thing that make them feel better, as you said, by reinforcing their critical thinking, reasoning, and that safe space that we will be able to see the positive sides of AI as well. So, thank you so much for being with us today. I'm sure a lot of parents will enjoy this discussion and learn from it, and learn how to talk with their kids' teacher about it to also help them in that journey. Thanks so much, Sarah.

SPEAKER_00 18:49

Thank you so much. I really enjoyed it.