Why AI Isn’t Killing Education: What Parents and Teachers Need to Remember
Last month, I sat at my scuffed kitchen table after school, half a bowl of cheddar Goldfish gone stale next to my stack of 10th grade To Kill a Mockingbird essays, and I found what every teacher is now conditioned to fear: a perfect essay. No run-on sentences, no misspelled character names, no messy tangents about how the movie was better than the book. Just a tight, polished argument that could have run in a high school literary journal. I ran it through an AI detector, and sure enough, it pinged 98% AI. The essay belonged to Javi, a quiet kid who plays third base on the junior varsity team and rarely turns in homework that isn’t smudged with dirt or riddled with spelling errors from his dyslexia. I called him in after school the next day, braced for a fight, for a denial. He walked in twisting his faded baseball team bracelet, and said, “I know it’s cheating. But I had all these things I wanted to say about Atticus, and every time I tried to write them down, they came out all messed up, and I knew I’d fail. I just wanted my argument to actually be read, not marked down for my spelling.”
That same night, my 13-year-old daughter was working on her science fair project on native bee populations at the other end of the table, and I glanced over to see she had ChatGPT open. I almost opened my mouth to tell her to close it, until I leaned in and heard her muttering at the screen. “No, that’s wrong,” she said, clicking back to her notes. “I counted 12 more squash bees in the pesticide plot than that, that’s not what my data says.” She was using AI to generate a template for her report, to organize her messy field notes, but she was arguing with it every step of the way, correcting its claims with her own work. That moment hit me harder than Javi’s confession had. All we talk about right now is AI killing education, how kids are cheating, how writing and critical thinking are going extinct. But what I saw that day wasn’t the end of learning. It was just a mirror held up to all the stuff we were doing wrong before AI ever showed up.
For decades, we’ve been grading the parts of learning that AI can do now. We mark papers down for bad grammar and spelling. We reward memorization of facts that any kid can look up in two seconds. We assign 5-paragraph essays that check a box for a rubric, instead of asking kids what they actually think about a book or an issue. We’ve been so focused on getting a clean, perfect final product that we forgot the point of the assignment is the thinking the kid does along the way. AI can write a perfect essay about Atticus Finch. But AI can’t have Javi’s experience of reading that book with his dad, who’s a public defender, and talking about how hard it is to do the right thing even when you know you’ll lose. AI can’t have my daughter’s experience of getting stung twice counting bees in our community garden, so she cares way more about her own data than whatever AI pulled from a generic study online.
After my talk with Javi, I changed how I grade essays in my classes, and it didn’t require any big expensive training or overhauls. Now, the first 50% of the grade isn’t the final essay. It’s a 5-minute voice note where the kid explains their argument to me, or a messy brainstorm page where they jot down all their thoughts before they start writing. If they use AI to clean up spelling or organize their notes, that’s fine, as long as they can show me their original thinking first. At home, I stopped telling my daughter she can’t use AI for homework. Instead, I ask her three specific questions any time I see it open on her laptop: What did AI get wrong that you know isn’t true? What idea did you add that AI didn’t come up with? How did this tool help you get to the part you care about faster? That takes 30 seconds, it doesn’t add any extra work, and it forces her to think critically about what the tool is giving her, instead of just copying and pasting.
Javi ended up redoing his essay after our talk. He recorded a 6-minute voice note of all his arguments about Atticus’s quiet flaws, used AI to turn that recording into a rough draft, then crossed out whole chunks of AI’s generic language and scribbled his own lines in the margins. The final draft had a few spelling errors, and it was messier than the original AI version, but it had teeth. It sounded like him. I gave him an A on the argument, and a B+ on the final draft, and he left my classroom grinning, saying he never thought he could get a good grade on an essay without hiding how hard writing is for him.
I still have bad nights. I still worry sometimes that I’m letting kids cut corners, that they won’t learn how to write a sentence on their own if they rely on AI too much. PTA meetings still devolve into panicked conversations about cheating, and I don’t have all the answers. But I remember when I was in high school, teachers panicked that calculators would kill math skills. Then when I was in college, everyone panicked that Google would kill research. Each time, we got so caught up fearing the new tool that we forgot the whole point of what we were doing. AI isn’t killing education. It’s just forcing us to stop teaching kids to do what a computer can do, and start focusing on the stuff that only they can bring. It’s messy, and we’re all still figuring it out, one essay and one science fair project at a time. But so far, that feels like a better way to teach than what we were doing before.
Please indicate: Thinking In Educating » Why AI Isn’t Killing Education: What Parents and Teachers Need to Remember