How I Used AI to Marie Kondo My Kid's Wardrobe and Toy Collection
The Starting Point
We had 47 t-shirts. She wore 8 of them.
I didn’t know we had 47 until I counted them. Somehow, over three years, 47 t-shirts had accumulated in our daughter’s drawer. Some were gifts. Some were from holidays. Some were “she might like this style” purchases. Some were things she wore once when she was three and we never threw out.
And she wore the same eight rotations. Every single day.
The other 39 sat there. Taking up space. Creating that low-level mental friction you get from a drawer that doesn’t close properly. Costing me mental energy every time I opened it trying to find the two shirts she actually liked.
That’s when I thought: what if I just photographed everything and let AI help me decide what to keep?
It sounds extreme. It was actually the opposite. It was the fastest, most thorough decluttering session I’ve had.
The Process (It Takes One Afternoon)
Phase 1: The Photo Session (30 minutes)
Lay every item on the bed. Photograph each one. You’ll have roughly 60-80 photos depending on volume.
It sounds tedious. It’s actually fast because you’re in a rhythm. Photo, move it to the “photographed” pile, repeat.
Tip: use good natural light. Take photos straight-on. AI will categorize these, so clarity matters.
Phase 2: The AI Sort (15 minutes)
This is where it gets interesting. Upload your photos to Claude or ChatGPT with this prompt:
“I’m decluttering my 7-year-old’s wardrobe. I’ve uploaded photos of [47] t-shirts. For each photo, tell me:
- Condition: Excellent / Good / Worn / Stained / Damaged
- Wear frequency (based on fit, fabric, style): High Use / Medium Use / Low Use / Unknown
- Size fit: Perfect / Slightly Small / Slightly Big / Too Small / Too Big
- Decision: Keep / Donate / Sell (on Carousell)
After you’ve analyzed all photos, summarize: - How many in each condition? - Total high-use items? - Any patterns (e.g., all the stained ones were from 3 years ago)? - Specific recommendations for what to keep/donate/sell?
Be honest. The goal isn’t to keep everything. It’s to keep what actually gets worn.”
Claude or ChatGPT will analyze the lot. You’ll get a clear breakdown. And here’s the thing: AI doesn’t have emotional attachment to the items. It won’t tell you to keep the shirt from the holiday trip three years ago just because of the memory.
Phase 3: The Kid Interview (15 minutes)
This is the part that matters most.
Show your child the AI’s list. Then sit down and go through the “keep” pile together. Ask:
- “Do you actually like this?”
- “When was the last time you wore it?”
- “Does it feel comfortable?”
Here’s what happens: your kid will have opinions you didn’t predict. They’ll say yes to things that seemed boring and no to things you thought they loved.
For my daughter, she wanted to keep all the bright colors and donate the pastels. I thought the pastels were “safe.” She hated them. She’d rather have fewer shirts she actively chooses.
This is the real insight: you’re not decluttering for them. You’re decluttering with them. Their opinion actually matters.
What The AI-Generated Inventory Revealed
After we’d sorted into Keep, Donate, and Sell piles, I did something else: I asked Claude to create a visual inventory.
The prompt: “Create a simple inventory list of what we’re keeping. Format it as:
WARDROBE INVENTORY 7-year-old Current Date: [date]
Shirt 1: [color] [style/print] [size] [when last worn] Shirt 2: ... etc.
Then summarize: - Total keepers - Color breakdown - Size range - Pattern (prints vs. plain) - Any gaps (e.g., ‘no warm layers for winter’)”
What you get is clarity. You know exactly what your kid owns. When it’s time to buy new items, you’re not guessing. You’re buying to fill actual gaps, not buying randomly because something looked nice.
This matters more than you think. For the next three months, I didn’t buy a single unnecessary item because I could see what we had.
The Donate/Sell Bit
The clothes you’re keeping go back in the drawer, organized by color or style (however your kid wants it).
The donate pile goes to a charity or hand-me-down group.
The sell pile (the ones in excellent condition but just not worn) goes on Carousell. You’ll actually get $2-5 per item, which adds up. My daughter and I got $40 from that pile. She wanted to keep half of it in her savings.
Here’s the surprising thing: she was more willing to let go of things once she realized someone else would wear them and she’d get money. It reframed the decision from “I’m losing something” to “I’m helping someone else and getting paid.”
The Toy Collection Version
After the wardrobe worked, I did the same with toys.
Same process: photograph, upload to Claude with a prompt asking to categorize by condition and “active use” (plays with this weekly vs. never touches it).
For toys, the decisions were sharper. Broken toys: donate. Never played with in six months: donate. Duplicates: keep the favorite, sell the other.
One thing AI caught that I wouldn’t have: we had five toy babies. She played with one. She had four others sitting in a bin because she’d been given them and we never dealt with it.
Once it was named (“you have five toy babies”), the decision was obvious.
What AI Can’t Do (And That’s Okay)
AI can’t navigate the emotional attachment part. Your kid will want to keep the toy from their best friend’s party in preschool, the one they never play with. That’s not a logic thing. That’s a memory thing. You keep it.
AI also can’t decide fit or practicality better than you. It can say “this has a stain,” but you know whether that stain bothers you or whether you’ll use it for painting anyway.
The AI’s job is to remove the decision paralysis. It gives you clarity so that the decisions you do make are based on real information, not vague feelings.
The Result
One afternoon. No kids crying about throwing out their favorite. No me standing in front of a drawer for 20 minutes trying to decide about the slightly-worn purple shirt.
Three piles. Clear decisions. A drawer that closes easily. An inventory I understand.
And the kicker: my daughter was involved. She learned that: - She doesn’t need 47 items to have a good wardrobe - Letting go of things you don’t use is okay - Someone else might want what you don’t use - Having fewer, better choices is actually more freedom
This is more valuable than the 39 extra t-shirts ever were.