I’m always looking for Ethical ways to use AI. Not only that, I’d like them to be fun. Things I can get my classes excited about. The technology is incredible, without question, but let’s not try to replace humanity. Let’s make humanity more interesting.

This idea started with something extremely unsexy: my kids’ report cards.

Middle school learning reports are packed with information but almost impossible to read at a glance. They always land at the busiest times of the year, so forgive me if I gloss over the pages of targets and teachers comments to focus on the grade.

Report cards, aren’t motivating beyond. They come too late to make a difference, but I thought what if they were more like the graphs the doctors have. The ones they give you to show your trends and what you can do to improve the outcomes. Data is motivating.

I tried an experiment.  I fed of the PDFs of my kids’ last three report cards into ChatGPT and asked it to:

Pull out the scores, turn them into simple graphs, and help me gamify the results so the kids could see their progress like a video game.

The result was a clear, non-shaming way to talk about school, what’s working and what’s not. Having three report periods made some things clear in a way it’s hard to argue.

Here’s what I did, why it worked, and how you can recreate it.

Step 1: No judgement. Treat the report cards like data

First, the ethics piece. These are my kids, not test subjects, so I set a few rules: Keep everything in a private ChatGPT session. No posting screenshots or raw PDFs anywhere. Strip out last names and teacher names. The model doesn’t need them. Use AI as a translator, not an authority. The scores and targets come from teachers; AI just spots patterns.

Then I grabbed the PDFs from the school portal and copy-pasted the sections with scores and the explanations.

My first prompt was something like:

“These are middle school learning reports for my child. The scores are 1–4, where 3 is on grade level. Please extract only the academic targets and scores from the last three grading periods.”

The model pulled everything into a clean table: grading period, subject, skill, score.

Step 2: Turn the year into a story with graphs

Once the scores were in a table, I asked the model to analyze them:

Average score by grading period (This is important for their phone usage guidelines)

Average score by subject across those periods

Then I asked for visuals.

“Create simple line graphs: one showing overall average by grading period, and one showing each subject’s average across the same periods.”

What came back was the overall story of their schooling.

Graph 1 – Overall line: one dot for each grading period, showing whether things were trending up, down, or basically flat.

Graph 2 – Subject lines: math, science, ELA, French, PE, etc. each got their own line across time.

For one kid, the overall line was gently rising. For the other, it was basically flat with a slight downward tilt as the content got harder. Neither graph screamed “disaster.” They just told the truth in a way that was easy to see.

More importantly, it gave us a story:

“Math took a hit when proportional reasoning showed up.”

“Science is quietly creeping up every trimester.”

“French has been a strength for two straight years.”

Instead of “I’m bad at school,” we could say, “This is the level where the game got tricky.

Step 3: Gamify the feedback so kids actually care,

Graphs are nice. Kids are not naturally moved by “average scores by content area.” So I asked ChatGPT to reframe everything in game language.

Prompt:  Create a slide show to show his work. Maybe like a gamification version of the graphics and explanations along with ways to improve.”

This was the result, minus details.
Title slide: “J’s Academic Adventure”

Slide with the overall XP bar.
Slide with subject-by-subject XP lines.
Slide of “unlocked skills” (their strengths).
Slide for “boss fights” (math, grammar, etc.).
Slide of new quests.

The tone was light, but the information was real. “You’re reading at a high level. That’s a huge advantage in every class.”

“Math isn’t broken; there’s one level that stood out as hard. Let’s take that on directly.”

“Science is trending up even when you feel like you’re just surviving labs.”

They saw themselves not as “good student/bad student,” but as players in a long game with different stats.

There are plenty of ways AI in education can go off the rails. This one worked for me because:

It respected the teacher’s work. I didn’t ask the model to re-grade anything. I used teacher scores and explanation as the source of truth.

It added clarity, not judgment. AI’s job was to surface patterns and offer language, not decide anyone’s future.

It gave the kids agency. The output was not “Here’s what you must do now.” It was “Here’s what the map looks like. What quest do you want to take first?”

And it modeled an important lesson about AI itself:
This isn’t a magic brain or a shortcut to avoid effort. It’s a tool that can help you see your own life more clearly and help you reach your goals.

How to try this with your own kids (or students)

If you want to steal this experiment, here’s the quick version:

-Grab at least three grading periods of report cards or progress reports.
-Upload or paste them into ChatGPT and explain the scoring system.
-Ask for a table: period, subject, skill/target, score.
-Ask for averages and graphs: overall and by subject.
-Have the model explain the graphs in kid-friendly language, then again in “game/XP” language.
-Work together to pick a few quests for the next grading period.

Everything else, slide decks, handouts, fancy visuals is optional.

The real win is sitting with your child, looking at the same picture of their learning, without having it be battle.

If AI can help us do more of that, then let the robots punch up some report cards.