AI Dad Lab
AI Dad Lab
Newsletter

← All issues

31 May 2026

🐿️ My 6-year-old is obsessed with squirrels. So we made one rob a Lidl.

Happy Sunday folks!

​

I found something this week that fixes the single biggest hurdle for parents wanting to try AI with their kids. Plus two experiments I ran, one with a holiday photo and one with a daft idea.

​

In this week's email:

​

  • 🐿️ Maria (6) loves squirrels and Lidl donuts, so we made an AI cartoon of a squirrel robbing a Lidl.
  • 🇨🇴 I turned a photo of a place we're about to climb in Colombia into a cinematic flythrough.
  • 🪄 The free library of 8,000+ prompts that fixes the "I don't know what to make" problem.

​

🐿️ My 6-year-old is obsessed with squirrels, so we made one rob a Lidl

​

Yesterday we found a whole family of squirrels in our local park, and Maria (6) has talked about nothing else since. She is also devoted to the donuts from the Lidl bakery. So this weekend we put both of her obsessions into one cartoon and made it together with AI.

​

First we got ChatGPT Images 2.0 to draw a storyboard, eight panels mapping out the whole story from a single prompt:

​

​

Then we fed that storyboard into an AI video tool (Higgsfield, running Seedance) and it animated the lot: a cheeky squirrel grabs a giant pink donut off the Lidl counter, a Garda gives chase, and it ends the only way it should, with a banana peel and a very undignified fall 😅.

​

​

👉 Watch the full chase: here.

​

The clever bit is the workflow: pick something your kid is mad about, get ChatGPT to draw the plan, then let the AI handle the movement and the comic timing.

​

🇨🇴 We're 10 days out from Colombia. So I flew through El Peñol.

​

As I showed last week, the kids have been counting down to our Colombia trip with a nightly infographic. One of the places we're going to climb when we get there is El Peñol, a giant 200-metre rock with ridiculous reservoir views from the top.

​

Here is the part that genuinely felt like magic. I took one still photo and drew a rough red line on it, just a squiggle for the path I wanted the camera to take: up off the water, past the bridge, around the rock. Then I handed it to Higgsfield (running a video model called Seedance). That scribble was the whole instruction.

​

​

It turned that line into this:

​

​

👉 Watch the full chase: here.

​

Honest verdict: the first seven seconds are jaw-dropping. The last few get woozy where I asked for two camera moves at once, and my source photo was a touch pixelated. But it built believable aerial footage of a real place from one still photo and a scribble. That was science fiction eighteen months ago.

​

🪄 The hardest part isn't the tools. It's the ideas.

​

Every parent I talk to says a version of the same thing. They have ChatGPT or Gemini open, the kids are keen, and then everyone just stares at the blank box. What do we actually make?

​

This week I found the fix. A free library called YouMind holds over 8,000 prompts for GPT Image 2, OpenAI's image tool, and every one shows the finished image next to the exact prompt that made it. The trick with kids is to not start from a blank page. Let your child pick an image they love, a comic strip, a trading card, a silly avatar, then tweak that prompt to be about them. Their team. Their pet. The story they are obsessed with this week.

​

To show you how it works, I customised two prompts in about two minutes each. First I took a "famous world cities as 3D maps" prompt and swapped in Colombia's four biggest cities, so the kids can learn the landmarks before we land:

​

​

And a "sports engineering poster" rebuilt around Désiré Doué and PSG lifting the Champions League at the Puskás Aréna last night:

​

​

👉 The free GPT Image 2 prompt library: here.

​

No signup, nothing to pay. Just scroll until something catches your kid’s eye, copy the prompt, and make it theirs. It’s the fastest way I’ve found to get from “I don’t know” to “we made that”.

​

That's the lot. If you do one thing this week, open the prompt library and let your kid pick something. I'd love to see what they make, just hit reply, I read every one.

​

Niall (AI Dad Lab)

​

---

P.S. If you'd rather start with prompts that already work, I put together 6 ChatGPT Images 2.0 activities for kids, character sheets, comic strips, sports posters and more, with a free PDF to print (link below).

​YouTube · Instagram · LinkedIn · Website

Was this forwarded to you? Subscribe here​

AI Dad Lab · aitoolsforkids.com
​ · ​

Liked this? Get the next issue.

One honest, parent-tested email a week. New issues arrive a week before they hit this archive, and the PDFs are subscriber-only.

No spam. Unsubscribe anytime.