UX Collective

We believe designers are thinkers as much as they are makers. https://linktr.ee/uxc

Follow publication

Contemplation in Blue: how to design large 3D sculptures with AI

I created a full-sized cubist head with DALL·E 3 and 3dmaker.ai and then used Blender to break it up into 3D-printable parts

Robert A. Gonsalves
UX Collective
Published in
12 min readJun 21, 2024

Three images, a screenshot of the model in Blender, a detail of a pluggable part, and the final sculpture
Screenshot of the Model in Blender (left), Detail of a Pluggable Part (center), The Final Sculpture (right), Images by Author

I wrote a Medium article a few months ago on using AI to create 3D printable objects. Due to the limitations of the printers I used, the largest piece I printed for that project was 7 inches tall.

For this article, I explored creating larger AI-generated sculptures with my new Bambu A1 Mini 3D printer. I devised a scheme to build larger creations using “plugs and holes” to join smaller parts together. The scheme is similar to the Lego concept but with bigger and more intricate parts. My final piece, “Contemplation in Blue,” stands 24 inches tall, 17 inches wide, and 9 inches deep.

In the sections below, I’ll explain my process for designing, segmenting, and printing large-scale sculptures.

Designing Sculptures with AI

To start things off, I logged into my ChatGPT Plus account, which integrates with OpenAI’s DALL·E 3 image generator. The Plus version of the service costs US$20 per month.

I used the new ChatGPT 4o model and gave it the following prompt.

Rob: Create an image of a modern, abstract, cubist bust of a man crafted from blue plastic, featuring a smooth surface with subtle indications that it was assembled using smaller pieces. Show a wide-angle view of the sculpture on a white pedestal.

The results are below.

The ChatGPT user interface for creating images with DALL·E 3
The generated image of a sculpture by DALL-E 3
ChatGPT User Interface for Creating Images with DALL·E 3 (left), The Image by DALL-E 3 (right), Images by Author

The first one I created looked great! The system did exactly what I asked. It even put a little description tag on the front of the pedestal, but it’s unreadable. Here’s what the response was:

ChatGPT: Here is the image of the modern, abstract, cubist bust of a man crafted from blue plastic. If you need any adjustments or additional details, please let me know!

That seemed helpful. When I clicked on the image, it showed a little info box with this text.

Create an account to read the full story.

The author made this story available to Medium members only.
If you’re new to Medium, create a new account to read this story on us.

Or, continue in mobile web

Already have an account? Sign in

Written by Robert A. Gonsalves

Robert A. Gonsalves is an artist, inventor, and engineer who writes about the creative uses of AI. Ask questions https://chat.openai.com/g/g-b1kqByRsT-robgonbot

Responses (5)

Write a response

Have you tried huggingface for free 2d to 3d. The state of the art is there and it's free, try the spaces

--

Thanks for sharing all the details of the process — the result is excellent. Great performance from a low cost printer! I really liked the animation of part assembly. Good job!!

--

Nice and a lot of work.

--