How I Created My Perfect Action Figure with Just a Selfie (NO ChatGPT!)

I created this image with ONE headshot and one prompt in Gemini with their latest Nano Banana model.

Below you can find both!

With a little bit back and forth I got this prompt - that nailed the environment and the fit.

Prompt:

"*Create a photorealistic, high-resolution image of a 1/7 scale commercialized figurine of me (face photos attached) placed on a stylish, modern desk of a high-end web designer. 
The Figurine: 
The figurine is of a man with a fit, athletic, and aesthetic build. 
His face should be realistic, well-defined, and accurately reflect the provided reference photos. All three depictions of the face (physical figurine, box illustration, and 3D model on screen) should be congruent and match the references. 
He is wearing a light pink hoodie with two visible white strings, loose-fitting stylish camouflage pants for a "chill" look, and obnoxious grey Nike Air sneakers with gum soles. 
The figurine is standing in a casual pose with hands in pockets on a simple, round, transparent acrylic base with no text. 
The physical figurine's overall appearance (face, attire) should precisely match the 3D model visible on the computer screen and the stylized illustration on the packaging box. 
The Environment: 
The desk is clean, minimalist, and organized, made of light-colored wood or a clean matte white material, matching the stylish aesthetic. 
In the background, a sleek, high-resolution iMac computer monitor is on. The screen clearly displays a 3D modeling software (like Blender or ZBrush) with the wireframe model of this exact figurine. The operating system shown on the screen must be macOS, not Windows. 
The iMac keyboard has white buttons and no visible cables. 
Next to the figurine is its premium packaging box, designed in the style of high-quality collectible figures. The box features a stylized, two-dimensional flat illustration of the character in the same outfit. The name "CKDML" is printed in a clean, modern font at the top, with the subtitle "The Automated Founder" directly below it. 
The surrounding environment around the iMac and desk should be very stylish, featuring green plants, a cool brick wall in the background, and amazing, soft, natural lighting. 
The image should have a shallow depth of field, keeping the figurine and the box in sharp focus while the background is slightly blurred.*"

Interestingly using multiple images of my beautiful ugly self resulted in a WORSE result! It could not get my face right with more data (or it screwed up the ienvironment):

Here's another good one with another prompt that's also left to look much more like the original:

All in all I have to say, I like these results a LOT more than the crap ChatGPT produced a few months ago. From image to text, it just nails everything with a little bit of tinkering. Yet it feels like no one's talking about it enough. We're entering the future of creation, the limits of turning an idea into reality just got a lot slimmer and I'm here for it.