Using ChatGPT to describe the style of images so you can reproduce them in DALL-E

I was sent this image by Helen Beetham in a discussion about potential logos/styles for a project we are working on:

I asked ChatGPT to analyse the image in order to describe the style. This is what it produced:

The image showcases a mid-century, retro-futuristic art style. Key characteristics include:

1. Bright, contrasting color palette with a mix of pastel and saturated tones.
2. Bold, simple shapes and lines.
3. Simplified details, especially on characters.
4. Subject matter reflects a blend of everyday life with advanced, imaginative technology.
5. Typography and layout reminiscent of vintage advertisements or comic panels.
6. Use of halftone patterns, which is typical of print media from the mid-20th century.

To reproduce this in DALL-E, you might provide keywords or descriptions like “retro-futuristic”, “mid-century modern”, “vintage

I then asked it to produce a logo titled ‘generative dialogues’ in this style. With the fairly major exception that DALL-E can’t spell, I was quite impressed by this process: