Latest / Campfire Guide to Parenting / Kids and AI: Never Smarter Than a Machine
Transcript
- Episode 5 Never Smarter than a Machine: A while back, the CEO of OpenAI was on a podcast and he was asked what it would be like to raise a kid in a world like this. Here's what he said. My kid is never gonna grow up being smarter than AI. Then he said, and that'll be natural. And then this. Of course, it's smarter than us. Of course, it can do things that we can't, but also who really cares? Now, I want to be fair to him. He might be right, but I've spent some thirty or more years. Around more than a hundred thousand kids, and I want to ask the question nobody asked him, Smarter than AI, at what? Because here's what I know. I have never watched a kid become confident by being right. And not once. Every confident kid I've ever known got that way by being wrong in front of people and then surviving it. So today I want to talk about what happens to a kid who never has to be wrong. This is Campfire Guide to Parenting and I'm Mike McDonald. Now, let me say something up front because I know how this kind of episode can sound, and I'm not here to tell you to throw the laptop in the lake. What I'm afraid of is quieter than that, because it's not that our kids will get a wrong answer from a machine. It's that they'll get the right answer four seconds way too early. Here's why this is on my mind. Common sense media surveyed about twelve hundred kids, nine to seven years old. And nearly a quarter of those kids said they would turn to AI for schoolwork help before they would ask a teacher or a counselor, or you, a parent. About one in ten said AI sometimes understands them better than most people do. For the kids who use it every day, that number is closer to one in five. And here's the part that I respect. The researchers said they can't tell which way it runs. Do lonely kids turn to AI more? Or does turning to AI more leave kids lonier? They just really don't know right now. Nobody does. We are the first generation of parents raising kids with this. And we've been running the biggest experiment now in history of childhood without a hypothesis. So I did what I do when I'm not sure. I went and I listened to the smartest people I could find. And they do not agree with each other. I find that kind of encouraging. Camp one is the optimist, okay? You've heard him, Sam Altman, who runs OpenAI. My kid is never going to grow up being smarter than AI. And that'll be natural. And he added this I think it's only weird for us in this transition time. I think that last line might be the most honest thing in the whole interview. It is weird for us, right? And I think the weirdness might be trying to tell us something. I'll also say this once, kindly. The man sells the product, right? And that doesn't make him wrong, it just means we keep our own judgment in this room. Camp two is my favorite because it's the builder who changed his mind. Sam Kahn, who founded Kahn Academy in 2023, on a TED stage, he promised giving every student on the planet an artificially intelligent but amazing personal tutor. That's a big promise. This past April, he told a reporter that About the tutor he built. He says, for a lot of students, it was a non-event. They just didn't use it much. And then he said this: AI is going to help, but I think our biggest lever is really investing in the human systems. Listen to that again. The man who built the tutor says the people matter more. That sits way much better for me personally. Camp three are the people who study what childhood is actually for. Alison Gapnick is a psychologist at Berkeley who has spent her career working on how kids learn. Her point is that we have these big brains because we get this long stretch of childhood where, in her words, we're protected. Protected so kids can wander around and open ended, no agenda, no great exploring. Which is a pretty good description of the best afternoons that I've ever spent seen at summer camp. And then there's Sherry Turkle. At MIT, who has spent her career studying how people relate to machines. She has a new book out, and her line is this What a chatbot does is offered offers pretend empathy. And about kids specifically, children can't learn trust from a device that only lies but doesn't know when it lies. Then there's Jonathan Haid, who of course wrote Anxious Generation. About AI companions and AI toys, he told parents, quote, with social media 15 years ago, we can say we didn't know. His point is that this time we don't get to say that because we're already seeing some of the harms. So, where does a regular parent land? Standing in the kitchen at 8:40 on a school night, homework is wide open, a chat bot is only one click away. Here's where I land. And you're allowed to totally disagree with me. I'm for our kids using it. I'm against our kids needing it to be capable. Those are not the same things. And this whole episode lives in the gap between them. So let me tell you about the kids that I've seen, right? And I'll leave out any kind of names, right? At camp, we often present kids with the opportunity to solve problems, challenges, like that's what we prefer to call them. Thousands of challenges over thirty-two years I've witnessed. A kid trying to figure out a way up a climbing wall, right? Struggling to find the path of least resistance. a group of kids strategizing on how to surmount a wall and get everyone over. A three-year-old presented with blocks, right, and asked to create a bridge. Each of these opportunities pushes its participants, stretches them to consider possibilities, and provided opportunities for struggle and failure, and then success. That kind of a learning process builds brains literally. It is no different than working a muscle. When we apply a little stress to our biceps, they got to grow back overnight in order to accomplish that same task with greater ease tomorrow. Here's what I've taken away from witnessing this type of learning. Real knowledge and learning come from the struggle, not the answer. Nobody can give a kid that. I can't. You can't. A chat bot certainly can't. It's only earned one hard minute at a time. Capable isn't something that we tell our kids they are. It's something that they find out about themselves. That's what really matters. I'm not saying any of this from a mountaintop. Believe me, there have been more than a few times that because of my lack of patience, the time involved, and just the desire to be done with homework, right? I've given my kid the answer. Or it's been because he's struggling. And I wanted to help him, right? Because I love him. That's what makes this so hard. We're not dodging their struggle because we're lazy. We're dodging it because watching them, watching them struggle hurts us, right? And now there's a button for that. So let me give you three tests you can run tonight. No rules, tests. You can use them in your own house with your own kid in your own words. And if your kid is five, none of this is about a chat bot yet. It's about whether you let them have. The 10 minutes. The tool question comes later. The habit, right, comes first. So test one, the struggle test. Did the tool replace the hard part or help the kid do the hard part? If the answer shows up before the confusion does, something got taken. Here's the rule that that I'd try. So 10 minutes alone with the hard thing first, then the tool. Never before. And when they do pick it up, Change the job. Don't ask for the answer. Ask it to quiz you. Ask it to find what you got wrong after you've already tried. Early research found that students working with a chatbot set up to withhold answers learned far more than students who couldn't just ask for them. Same technology, different job. All right. Test two. The first call test. When something goes wrong, who does your kid reach for first? Remember, nearly a quarter said AI before parent. And that's for schoolwork, right? Schoolwork is the easy stuff. I care about the hard stuff, right? Then I said something awful stuff. Then I'm scared stuff. The nobody sat with me at lunch stuff. Here's my read. And it's mine, not a study's. Kids take the hard thing to whoever handled that last hard thing well. First time your kid tells you something difficult and you flinch, you've taught them to take it to the next person or the next thing, right? So be steady. Say, thank you for telling me that before you say anything else. All right, last one, test three. The vulnerability test. Is this a place where my kid practices relationships or hides from them? Turkle again, right? She says, It turns out that you can't have intimacy or friendship without vulnerability. The chat bot never gets tired of you. It never has a bad day. It never says no. Real friends do all three. And that friction isn't a bug, it's the whole curriculum of friendship. I talked about this in the friendship episode, right? Kids are wired to connect. They just need the chance. So if your kid is lonely and the chat bot is the only thing that feels easy, the answer isn't to rip it away. It's to fill the space, it's filling the place. With a person. Now, some of you are listening and thinking, Mike, my kid has a learning difference, and this is the first thing that's ever been patient with him. Or we can't afford a tutor, and this is the only help out there. and at the table at nine o'clock at night, I hear you. Use it. I'm not asking you to put it down. I'm asking one question every time. What does it replace? If the answer is a wall my kid couldn't get over alone, that's a bridge. If the answer is the 10 minutes where my kid wouldn't have figured it out, that's a shortcut. Same tool, different question. Now, the part nobody wants to hear, including me, we're all doing it too, right? It's two in the morning. You've got a parenting question and instead of calling the friend the next day who's been through it, you just type it into the box, the answer pops up. Calm and confident, and it's never once awkward. We ask it what to text our teenager. We ask it to write the email to the teacher. We ask it to make up the bedtime story we used to tell badly in the dark. Each one of those is small. Each one is also a vote. Our kids are watching us decide that not knowing is a problem to be solved in those four seconds. The test for us is the same as it is for them. Did it replace the hard part? And the hard part of parenting was never having the right words. It's being in the room, not knowing, and staying anyway. A kid who watches a parent sit in not knowing learns that not knowing is survivable. That might be the most important thing we can teach them right now. And it's the one thing we can ask a machine to model for us. So before you go, here's what I would like all of us to contemplate on our own. When your kid is stuck or scared or lonely, who do they call first? And what did you do this week to stay that call? Thanks for hanging out with me. And this is Campfire Guide to Parenting. I'm Mike McDonald, and I'll see you around the campfire next time.