Kill with a Borrowed Sword
Use AI as Your Child's Training Partner, Not Their Replacement
Sun Tzu used an enemy's resources against him. I am doing the same thing with AI in my house. The tool is already here. My daughters will either learn to use it as a sparring partner, or they will let it do their thinking for them. The sword is in your house too. The question is who holds it.
My 13-year-old came to me last month with a history essay. I watched her use a chatbot to argue against her own thesis until she found the holes in her reasoning. She was not asking it to write for her. She was using it like a debate partner who never got tired. I have seen versions of this with all three daughters now. Some productive, some not. It made me pay attention to how other families handle the same tension. The scenario below is a composite, but it is close to what I have watched play out.
Brandon teaches sophomore English in a suburban district outside Phoenix. Last October, he caught three students submitting essays that were clearly AI-generated. Two denied it outright. The third kid, Ethan, admitted it immediately and asked if Brandon wanted to see the eighteen drafts he had gone through with the model to get to something worth turning in.
Brandon sat down with Ethan after school. The kid pulled up his ChatGPT history. Draft one was garbage. Generic, flat, exactly what you would expect from a prompt like 'write me an essay on Gatsby.' But then Ethan had asked the AI why the introduction was weak. He pushed it to suggest three different opening hooks. He asked it to critique his thesis for logical gaps. By draft seven, Ethan was arguing with the model. He told it that its interpretation of Gatsby's shirts scene was too literal.
Brandon read through the final essay. It was better than anything Ethan had written all semester. Not because AI wrote it. Because AI had forced Ethan to defend every claim, clarify every transition, and articulate what he actually believed about the text. The tool had not replaced his thinking. It had sharpened it.
Brandon changed his policy that week. No AI was off the table. But every student had to submit their conversation history alongside their final work. The goal was not the essay. The goal was the thinking. If AI could be the sparring partner that forced better thinking, Brandon was going to use it.
Within three months, the quality of student writing in his class improved measurably. Not because they were cheating more effectively. Because they were learning to use the tool the way a boxer uses a heavy bag. You do not let the bag do the work. You use it to train the punch.
I did not understand this right away. My first instinct when I saw my 9-year-old asking a chatbot to explain her math homework was to tell her to figure it out herself. Like that was the only way learning happened. It took me a few weeks to realize I was confusing the method with the outcome. The parents who win the next decade will not be the ones who keep AI away from their children. They will be the ones who teach their children to use AI as a mirror, a critic, and a sparring partner. This is not about becoming an AI expert. This is about understanding that every tool that threatens to replace human capability can also be used to develop it.
A 2024 study from Stanford's Graduate School of Education found that students who used AI tools to critique their own work before submission demonstrated 34 percent stronger critical thinking skills on independent assessments than students who either avoided AI entirely or used it only for content generation.[1] The gap was not about access. It was about intent. Students who treated AI as a co-pilot learned to second-guess their assumptions. Students who treated it as a vending machine learned nothing.
Your child will have access to AI whether you approve or not. The average teenager already uses generative AI 4.7 times per week, according to a 2024 Pew Research study. That number is climbing.[2] The question is not whether they use it. The question is whether they use it to avoid thinking or to improve thinking. That difference will determine whether they enter adulthood as a tool user or a tool dependent.
The borrowed sword stratagem works because it turns the opponent's weapon into your advantage. AI is already in your child's hands. The strategic parent does not confiscate it. They teach the child to sharpen themselves against it.
Tool leverage is not about using AI to get things done faster. It is about using AI to force your child into deeper engagement with their own reasoning. When your ten-year-old asks ChatGPT to explain photosynthesis, the win is not the explanation. The win is what happens next. Do they stop there, or do they ask three follow-up questions that expose gaps in the model's answer? Do they take the explanation and try to teach it back to you in their own words? Do they test whether the explanation holds up when applied to a different plant?
A 2024 analysis from MIT's Center for Collective Intelligence tracked 1,200 high school students over six months as they integrated AI tools into their study routines.[3] The students who showed the strongest academic gains were not the ones who used AI most frequently. They were the ones who used AI to generate questions, not answers. They used the tool to simulate a debate opponent, a skeptical peer reviewer, a confused student who needed concepts explained simply. They treated the tool as resistance training for their own thinking.
This is the shift most parents miss. They see AI as a shortcut, and they are not wrong to worry. But the borrowed sword strategy flips the frame. Instead of asking 'how do I keep my child from using AI to avoid effort,' the strategic parent asks 'how do I teach my child to use AI to increase effort in the places that matter.' The tool becomes the coach that never gets tired, never runs out of patience, and never lets a weak argument slide.
Use another's strength as your weapon. Attack using the strength of another when you are on the offensive.
AI has infinite patience and no ego. Teach your child to exploit both.
AI can generate an essay, solve a calculus problem, write code, summarize a book. It cannot develop your child's ability to recognize when an answer is plausible but wrong. It cannot teach them to notice the moment when their own reasoning breaks down. It cannot force them to defend a position they do not yet fully understand. Those capabilities require a human being who is willing to be challenged, corrected, and confused. The fastest way to develop them is to put your child in conversation with a tool that will challenge them without judgment.
A 2024 report from the World Economic Forum identified 'critical evaluation under conditions of uncertainty' as the single most in-demand cognitive skill for workers entering the labor market in 2030 and beyond.[4] Not coding. Not data analysis. The ability to take an ambiguous situation, stress-test multiple interpretations, and defend a course of action even when the data is incomplete. That skill is not built by consuming information. It is built by arguing with it.
When your fourteen-year-old uses AI to write a history essay and you ask them to explain why the AI's thesis is either stronger or weaker than their original idea, you are training critical evaluation. When your nine-year-old uses AI to help with math homework and you make them explain the steps back to you without looking at the screen, you are training retention and articulation. The tool does not do the learning. It creates the conditions for learning to happen.
This is why the borrowed sword works. You are not fighting AI. You are not banning it. You are using its existence to force your child into the kind of cognitive labor that builds the skills AI cannot replicate. The sword cuts both ways, but only if you teach your child how to hold it.
I made this mistake with my oldest daughter for about two months before I caught myself. I saw her pulling up a chatbot to work through a science project and immediately assumed she was taking a shortcut. I almost shut it down before I asked her to show me what she was actually doing. Turns out she was using it to test whether her hypothesis made sense before she started gathering data. That is exactly what I would have done if I had access to that kind of feedback loop when I was her age. The mistake is thinking that access equals dependency. Parents see their child using ChatGPT and assume the child is outsourcing their thinking. Sometimes that is true. But access to a tool does not determine dependency. How the tool is used determines dependency. A calculator makes you dependent only if you never learn the underlying math. But if you use the calculator to check your work, catch errors, and test edge cases, it makes you faster and more confident. The same logic applies to AI.
The second mistake is waiting for schools to solve this. A 2024 survey from the OECD found that 72 percent of secondary school teachers have no formal training in how to integrate AI tools into pedagogy, and 61 percent of school districts have no coherent policy beyond 'do not cheat.'[5] Your child's school is not equipped to teach tool leverage. You are. The window is short. By the time most institutions catch up, your child will have already formed their habits.
The third mistake is assuming that using AI well requires technical expertise. It does not. It requires the ability to ask good questions, recognize weak answers, and insist on clarity. Those are parenting skills, not engineering skills. If you can help your child think through a tough decision, you can teach them to use AI as a thinking partner. The borrowed sword strategy is not about mastering the tool. It is about teaching your child to stay in control of their own reasoning while the tool does the heavy lifting.
What This Looks Like in Practice
Two ways to teach tool leverage this weekend—no technical expertise required.
- Pick one task your child does with AI this week. Homework, research, anything. Before they finish, make them ask the AI: What is the weakest part of this answer? Have them defend or revise based on what the AI says.
- Teach the follow-up reflex. Every time your child gets an AI answer, they must ask one clarifying question before moving on. No exceptions. The question can be simple. Why? How do you know? What if the opposite were true? But it must happen every time.
- Make the AI argue against your child. Pick a topic they care about. Sports, games, school policy, anything. Have them state their position, then make them ask the AI to argue the opposite as convincingly as possible. Their job is to defend their original view or admit the AI has a point. Either outcome is a win.
The research behind
this framework, weekly.
Raised Nimble is a free weekly newsletter that translates AI and future-of-work research into plain-English guidance for parents. Each issue is one concrete idea you can act on before the weekend. No jargon. No fluff.
No spam. Unsubscribe anytime.