The Problem Most Parents Are Seeing
Your child opens ChatGPT or another AI tool and types, "Write an essay about the French Revolution."
Ten seconds later, they have 500 words. Copy, paste, homework done!
But what did they actually learn? Nothing.
What skills did they build? None.
And when they're asked about the same topic later in class, they can't answer, because they never truly learned it in the first place.
Our job as parents isn't to ban AI. It's to teach our kids how to use it in a way that makes them smarter, not dependent.
The 80% Rule: A Framework
Your child must complete 80% of any work themselves, a full first attempt, before AI enters the picture.
And when AI does enter, its job is to critique, not complete.
They do the thinking. AI points out what's weak. They fix it themselves.
The Core Principle
AI should amplify thinking, never replace it.
What Counts as 80%?
Your child has done 80% when all three are true:
- They have a complete first draft. Not an outline. Not bullet points. Something they could submit if AI didn't exist.
- They can explain their reasoning. If you ask "why did you write it this way?" they have an answer.
- They hit a wall and pushed through. They encountered something hard and made a decision, even if it's wrong.
What doesn't count:
- "Let me just see what AI says first."
- "I wrote the introduction, can AI do the rest?"
- "AI, make this better."
What It Looks Like in Real Life
My 13-year-old had an essay due on the French Revolution. His first draft was rough, weak thesis, vague conclusion. But it was his, a 400 word genuine attempt.
After finishing, he asked AI one question:
"What's the weakest part of this essay? Don't fix it, just tell me what's wrong."
AI pointed out the gaps and he went back and revised it himself.
The final essay wasn't perfect, but he actually understood the topic. That's the difference.
Why This Works (The Science)
Researchers call it "desirable difficulty." When something is a little hard, we learn more from it.
When things are too easy, they don't stick. But when we struggle, make mistakes and then figure it out, the learning lasts.
AI that does the work removes the difficulty.
AI that critiques the work preserves it, while adding personalized feedback.
Think of it like weightlifting. Having someone else lift the weights doesn't make you stronger. But having a coach watch your form and suggest corrections? That accelerates real growth.
Age-Specific Implementation
Ages 6-9: No AI for Their Work
They're building foundational skills. Handwriting, basic reasoning, imagination. AI can wait.
Ages 10-12: The 80% Rule (Training Wheels)
This is when to introduce the framework. But don't just announce it, teach it.
First few assignments:
- Sit with them while they do their 80%
- Help them craft the critique prompts
- Watch them process the feedback and revise
- Gradually step back as they get it
Review their AI usage weekly for the first month. Then move to spot-checks.
Ages 13+: The 80% Rule (Independent)
By now, they should run the framework themselves. Your role shifts from supervisor to occasional auditor.
Check-in frequency: Random spot checks. Maybe once every week or two.
What to watch for:
- Are their first drafts getting shorter over time? (Bad sign)
- Can they discuss their work intelligently? (Good sign)
- Do they seem dependent or capable? (Trust your gut)
When They Push Back
"This is stupid. Everyone else just uses AI."
"I know it feels like extra work. But here's the thing, everyone's going to have AI. The kids who can think without it will stand out. I'm giving you an advantage, not taking one away."
"I'm stuck. I can't even get to 80%."
"Stuck is fine. Stuck is where learning happens. Let's figure out what's blocking you, but 'stuck' doesn't mean AI does it. It means we work through it."
"But I literally don't know where to start."
Ask them: "What's one thing you know about this topic?" Start there. If they truly know nothing, they can research (including asking AI to explain the topic), but they still write the work themselves.
The Bottom Line
AI is powerful, but your child's ability to think is more important.
The struggle is where learning happens.
The 80% Rule protects that struggle, while still teaching kids how to use AI well.
This is one framework.
The book has the complete system.
AI companions your teen might already be using. Deepfake protection. Conversation scripts for every pushback. Activities organized by thinking skill. Age-by-age playbook from 6 to 16.
See The Kind Machine →The Fridge Checklist
Print this. Put it where they do homework.
Before Asking AI Anything:
Did I finish my own version first?
A complete draft I could submit, not an outline.
Am I asking for critique, not answers?
"What's weak?" not "Fix this for me."
Will I improve it myself?
Using my own thinking and my own words.
Frequently Asked Questions
What is the 80% Rule for AI?
The 80% Rule is a framework where kids complete 80% of their work themselves before AI enters. When AI does enter, it critiques the work, it doesn't complete it. The child then improves it themselves.
What age should kids start using ChatGPT?
Ages 6-9: No AI for their work. Ages 10-12: Introduce the 80% Rule with parental supervision. Ages 13+: Independent use with occasional check-ins.
How do I know if my child actually did 80% themselves?
Ask them to explain their work. If they can walk through their reasoning, they did the thinking. If they stumble, they probably outsourced too much. Also ask to see their AI prompts.
Should kids use ChatGPT for homework?
Yes, but only after they've done the work themselves first. AI should critique their draft, not create it. The 80% Rule ensures they do the thinking before AI enters.
What's the difference between using AI for research vs cheating?
Using AI to understand a concept is learning. Copying AI's explanation into your assignment is dependency. Research with AI is fine. Writing with AI is the problem.