How to make a ‘Barbie Box’ AI action figure with ChatGPT

Social media viral trends come and go and the latest is people showing off action figures of themselves as if they’re real merchandise. It’s called the ‘Barbie Box’ trend and you can make your own.

These, as well as the Studio Ghibli style images, are being shared across the likes of Instagram and LinkedIn, and it’s easy to see why – they’re funny and eye-catching and you can make one of yourself quickly and for free.

These AI action figures look much like a Barbie or Action Man from a toy shop and are possible thanks to a recent update to ChatGPT’s image model, but other image generators may also be capable.

Follow the steps below to make your own AI action figure, but think about how you’re using it first. It’s a bit of fun, but please consider using a real artist for commercial use as well as the energy it consumes.

1. Open ChatGPT and log in
2. In the prompt box, type “Make an action figure of me”
3. Give ChatGPT more information on what you’d like it to look like as per its reply (the more specific the better) and upload an image of yourself
4. Hit enter and wait (a while) for it to generate the image
5. From here, you can ask for adjustments until you’re happy with how it looks

In my example, I gave it the following information as well as a photo of myself but then refined it over six different versions until pretty good. You can see a couple of iterations below.

• Title: Reviews Editor
• Catchphrase: ‘Too Many Phones’
• Clothing: Hipster tech clothes
• Pose: On a phone call
• Accessories: A camera, laptop, backpack
• Packaging: Retro packaging, like a collectable