What Teachers Are Actually Telling Kids About Their Future With AI
Last month I sat in on my 10th grader Leo’s parent-teacher conference, the fluorescent lights hummed off-tone, and Ms. Henderson, his graphic arts teacher, had a chipped ceramic coffee mug that read “world’s okayest art teacher” on her desk, with a stack of crumpled student portfolios leaning against the leg. Leo loitered in the doorframe scrolling his phone, waiting for us to finish, and I asked the question I’ve turned over every night for a year now: what do you actually tell kids about their future, with all this AI stuff going around?
Ms. Henderson leaned back, crossed her arms, and admitted she spent the first half of last year panicking. She caught three kids turning in entirely AI-generated posters for the fall play, no edits, no original input, and stayed up one night reading doomsday thinkpieces claiming all creative jobs would be gone in 10 years. Then she stopped panicking and started figuring out what kids actually need to hear, instead of repeating the same panic she saw online.
The first thing she tells every student, she said, is to stop asking “will AI take my job.” That question just makes you freeze, waiting for someone to tell you if your dream is worth chasing. Instead, she makes them ask: “what can I give this work that AI can’t?” She used Leo’s recent homecoming poster as an example. Leo had used MidJourney to generate 15 base designs, all crisp, all pretty, all technically good. But AI didn’t know our student body has a running three-year joke about the moldy cheese pizza in the cafeteria, so Leo slipped a tiny cartoon slice of pizza in the corner that only other kids would catch. AI didn’t know the gym’s banner rack only fits a 3:1 ratio, so Leo cut and rearranged the whole design to fit. AI didn’t know the student council president hates yellow with a passion, so Leo swapped the background to soft navy. None of that was advanced technical skill, but none of it was something AI could do on its own—because AI doesn’t go to our school, doesn’t know our inside jokes, doesn’t care if the student council president is happy with the final product.
I left that conference thinking about how I used to snap at Leo when he admitted he used AI for homework. I’d lecture him about learning to do it on his own, worried he’d never build real skills. Now I try a different routine, something I picked up from what Ms. Henderson shared. When Leo says he used ChatGPT to outline his history paper or MidJourney for a personal project, I don’t get mad. I just ask two specific questions: “what did AI get wrong that you fixed?” and “what did you add that’s totally yours?” It takes 30 seconds, it doesn’t start a fight, and it gets him thinking about the part he plays, even when he uses a tool to speed up the work.
I ran into Leo’s computer science teacher at a soccer game a week later, and he said the same shift is happening in his classroom. He used to spend half the semester teaching kids how to write basic loops and memorize syntax. Now AI can crank out that base code in 10 seconds. So he reworked his curriculum to spend that time teaching kids what AI can’t do: how to read the code AI spit out and spot errors, how to tweak it to fit what the user actually needs (not just what the AI thinks you need), how to test for gaps that AI didn’t think to cover. He tells his students the best coders today aren’t the people who can memorize every line of code—they’re the people who can tell AI exactly what they want, then fix the mess it makes.
Most of the advice educators are sharing right now boils down to small, concrete shifts, not big overhauls of how we parent or teach. Ms. Henderson makes all her students write down their AI prompts before they generate anything, then write down three rounds of edits they made to the first result. At home, that translates to a quick, low-effort check-in: if your kid uses AI for a school project, just ask them to show you the first output and the final result, and talk through what changed. It builds the habit of thinking through what you need, instead of just taking what AI gives you.
I grew up with my dad telling me “learn a skill, stick with it, and you’ll retire from the same company.” That worked for him—he was a mechanic at the same shop for 32 years. But that’s not the advice teachers are giving kids now. They don’t tell kids to pick one “AI-proof” career and stick with it, because no one knows what will be safe in 10 years. Instead, they tell kids to pick something they’re curious enough about that they’ll want to keep learning it as it changes. When I told Ms. Henderson I worried Leo picking graphic design was a bad bet because of AI, she just shrugged and said “he spends three hours a day messing around with designs for fun, even when it’s not for school. That’s the best predictor he’ll be fine. People who love what they do will adapt to the tools.”
I still worry some nights. I’ll see a clickbait headline about mass automation and lie awake wondering if we’re doing enough to prepare him. But yesterday I saw the birthday card Leo made for his grandma. He used AI to generate a picture of her old beagle Max, who died 10 years ago. AI got the spots wrong, missed the little nick Max had on his left ear, and made his tail too long. So Leo got out his colored pencils, redrew the whole face, added the nick and the correct spots, and wrote on the inside “remember when he ate your entire Christmas ham off the counter?” That’s the part that made grandma cry. AI couldn’t remember that. AI couldn’t care that it mattered. That’s what the teachers are telling kids: the stuff that matters is still ours. I don’t know exactly what the future will look like. No one does. But so far, this quiet, grounded advice makes a lot more sense than all the panic.
Please indicate: Thinking In Educating » What Teachers Are Actually Telling Kids About Their Future With AI