Should I teach my kids to use AI?

This week, for the respective editions of their newsletters, Kids Today and User Friendly, senior correspondents Anna North and Adam Clark Estes have a chat about something that’s occupying both of them as journalists and as parents of young kids: How AI will shape the lives of the next generation? Anna North: Hi Adam! Excited to chat about AI and kids! In addition to reporting on kids for work, I also have a 7-year-old (who is home sick today and watching Amphibia right now, so might interrupt), and a 2-year-old (who is not currently home sick, yay!). Adam Clark Estes: Hey Anna! My kid is not yet 2, but as they say, the days are long, but the years are short. I feel like I’ll be meeting her AI friends in no time. Anna: Ha! So speaking of which, should we kick this off by talking a little bit about our hopes and fears for our kids growing up in the age of AI? I do feel like there’s a lot to be hopeful about: I’m excited that my kids will grow up in a time when we can use AI to develop new drugs and decode elephant language (although maybe some of those AI tools are more properly called machine learning?). My biggest fear is that my kids won’t see the point of learning certain skills, like writing and math, because AI can do those tasks for them. I certainly worry about AI taking jobs, too, but that’s a ways away for my kids. Adam: I’m actually not that worried about my kid having AI friends. She probably will! I just hope that she doesn’t spend too much time with them or too much time with AI-powered technology in general. I hope we enter this new AI era together and learn how these tools can make our lives better, richer, and more interesting. It reminds me of how the internet came of age around the same time I did, but my parents seemed too afraid to figure it out at the time. I hope I can be brave. Anna: My dad was actually an internet early adopter, and so we did enter that era together, which was sweet. We used to go to Doctor Who chatrooms — I remember once listing “our” age as “43 and 12.” That was nice for me because 1) I learned to use a computer and 2) I learned to be curious and not fearful around new technologies. Except I am a little fearful around AI! Maybe I have to channel that spirit of curiosity a little more when I’m with my kids. A friend of mine was telling me her district is rolling out AI tools potentially in kindergarten, so it starts young. Adam: And I keep seeing ads for AI-equipped stuffed animals. AI is still such a buzzword, but I do think we’re starting to see how it really will take over technology as we know it. I always say that it’s just the next iteration of the software that’s already in everything. So, of course, it’s going to be in classrooms — not just for cheating purposes, either. Anna: Cheating using AI is less of an issue for little kids because they’re not doing much of their work on computers yet. The AI tools I’ve seen proposed for K-6 are like this AI-powered reading coach, which seems less vulnerable to cheating than, say, ChatGPT. I’ve definitely heard people say that AI could potentially offer more personalized tutoring one day, which could be positive since individual instruction is really beneficial, but public school classes are often too large to offer much of it. That idea doesn’t freak me out too much, although, of course, there are questions around how reliable and accurate AI tools are. My older kid is in first grade, and at his school, there’s a lot of talk about the role of AI in math. The argument I’ve heard is very similar to what we were taught about calculators, honestly: that these tools will be able to do a lot of simple operations for us, so kids should be able to think intuitively about numbers. Adam: I’m also curious about how AI simply changes how schools work. Like, if AI makes it too easy to cheat when essay-writing, what will teachers assign instead? And what will they grade? I don’t really think we’ll return to oral exams, but we will have to reconsider what education looks like in a world that runs on AI. Anna: I do think there’s a shift to more in-class tests and assignments, which can cause its own stress. I also wonder if there’s just going to be a shift toward developing a different set of skills, if writing just becomes less important. We already see kids and young people consuming less text — I wouldn’t be shocked if there was more emphasis in the coming years on oral presentation or audio and visual production skills. That bums me out as a writer, but maybe it’s what Plato would’ve wanted? I do try to remember that people have always been skeptical of new technology (even written language), and some of the anxieties we’re experiencing now are thousands of years old. Adam: And some look like instant replays of anxieties from just a decade or two ago. Every time a new technology or medium comes along, there’s a collective freakout that it’s destroying the youth. This is just as true for social media as it was for TV, video games, comic books, and even radio. I think this will be true of AI in social settings, too. It’s already possible to have an AI friend through apps like Replika or Kindroid. I wasn’t kidding about the AI stuffed animals, either. There’s one called Grok that’s designed — by Grimes of all people — for ages 3 and up. Kids can ask it questions, and the AI will tell them things, like a futuristic Teddy Ruxpin, except Teddy Ruxpin had a cassette tape in his belly that parents could listen to and know what the stuffed bear would say. I have no idea what AI will be telling our kids! I’ve seen people call this the end of the imaginary friend. I actually think it’s just the beginning of something new. What that is, I can’t imagine. At least not yet. Anna: It’s funny to me because Teddy Ruxpin was famously creepy! The social/play aspect of AI is super-interesting. I’m not worried about AI killing imaginary friends — kids will make friends with a can of tomato paste, and I don’t think you can destroy their social creativity that easily. My biggest concern around AI friends right now is safety — there are already lawsuits alleging that chatbots nudged kids toward violence or self-harm. Do you know what guardrails there are in place? Adam: My general feeling about guardrails is that, no matter how many there are, technology finds a way to leap over them. YouTube, for instance, has long struggled with how to make sure parents can steer their children to safe, age-appropriate content, but kids inevitably find themselves sucked down a rabbit hole of uncanny algorithmically generated videos. Throwing an infinite supply of AI-generated content into the mix won’t help, so I think parents will have to be vigilant about triple-checking what their kids are watching or playing with. There was a Pew study earlier this year that said about a quarter of all teens had tried ChatGPT for schoolwork. That number had doubled in a year. So if you assume that guardrails aren’t there or won’t work and that kids are going to try some kind of AI tool eventually, where does that leave parents? To be honest, I think we should all do what your dad did with you: Hang out in the proverbial chatrooms together. Talk to the chatbots together. Play with the AI toys together. Learn about this new technology along with your kids and help them learn when to put them away. Anna: I’m sure my dad will appreciate this endorsement of his parenting! You’re not alone, though. Andrew Przybylski, an Oxford professor I think we’ve talked about who studies phones and kids, talks about introducing his children to smartphones the way you would teach a kid to ride a bike: It’s a tool. It has hazards but also uses. It can be fun, and it’s a basic part of life. Maybe the same is true of AI? This conversation is sort of making me think I need to use AI more with my kids, which is not where I expected to end up. Adam: It’s important to point out that we’re mostly talking about generative AI here, and chatbots. There are also image and video generators. These all have obvious applications for kids in schools, for cheating and learning. But we haven’t even gotten into what the next generation of AI will impact our kids’ lives — things like AI agents that can use computers themselves or the much feared artificial general intelligence that can theoretically do anything. That future is a lot harder for me to comprehend right now. Anna: Yeah, I think there’s a lot about the next 10 or 15 years, both in AI and just in our kids’ lives generally, that’s hard to wrap one’s head around. My husband and I always joke about our children going to college on the moon, but I think it’s just a way of expressing the uncertainty that’s always there when you try to project too far out.

This moment demands more than headlines

Should I teach my kids to use AI?