They can prompt themselves:
my prompt to the ai: write a prompt for a fine art photo
response:
Here's an example of a well-crafted prompt for generating a fine art photograph using an AI image generator (like Midjourney, Stable Diffusion, or DALL-E). This one is designed for a surreal, conceptual black-and-white fine art portrait, inspired by classic fine art photography styles:
Prompt:"A haunting surreal fine art portrait of a solitary ethereal woman emerging from dense fog in an abandoned Victorian greenhouse, her face partially obscured by delicate floating feathers, eyes closed in quiet contemplation, dramatic chiaroscuro lighting with deep shadows and high contrast, monochrome black and white, intricate details in textures of glass and foliage, emotional depth conveying isolation and mystery, shot on large format film with shallow depth of field, in the style of Francesca Woodman and Sally Mann, hyper-detailed, 8k resolution, masterpiece fine art photography"
Tips for customizing or creating your own:
- Subject: Describe the main element clearly (e.g., person, landscape, object).
- Style/Mood: Add "fine art photography," "surreal," "conceptual," or reference photographers (e.g., Cindy Sherman, Gregory Crewdson).
- Technical details: Include lighting (e.g., golden hour, low key), camera/film (e.g., Hasselblad, Kodak Portra), depth of field, composition.
- Quality boosters: End with "highly detailed, masterpiece, 8k, sharp focus."
I fed the above prompt into nanobanana:
View attachment 415193
My only action was to ask for a prompt and to submit the prompt it wrote. This was a very crude example but not hard to imagine this being 100x more capable a year from now.