Embedded Files
Finding The Right Story
And Tone of Voice of AI Driven Photo Art
With "Wasted Chances" series, I wanted to create images that with the first impression would seem as beautiful, natural photography, but contain a deeper layer of conversation. With the series, I wanted to ask the question we don't want to hear: What does our lifestyle look like from the very birthplace of human kind? (Where our e-waste largely ends up.)
First, I prompted AI to create an iconic, natural photography of giraffe reaching to eat from a tree. The prompt included the composition and set up, the message/story what the image should represent, lighting, atmosphere, camera angle and settings.
I prompted the e-waste as wires on the tree, and we would see the giraffe actually nibbling the wires instead of the leaves of the tree.
I wanted the photo to represent the giraffe really reaching and the photo to show the wires subtly, so the viewer would have the joy (or the sadness, in this case) to realize what is actually happening in the image.
The image required several reprompts as well as editing afterwards.
I wanted the format to be vertical, to emphasize the nature of the animal and to really get a striking composition. This image, the colors and the lighting are what I was looking for (I was especially happy with the light shining through the leaves), but the composition is off. Even though the depth and the branches are just right, unfortunately the giraffe is way too casual and the wires are a little too subtle.
I also prompted two giraffes, the idea being they could emphasize each other's composition and give a subtle story.
This time, I prompted the light to be the golden hour, and as backlight, to tell the story from the savannah.
This photo has the light I was looking for, but unfortunately leaves the giraffes too much in the shade for this story. The depth and composition might work otherwise, but the composition of the e-waste does not work.
I prompted a horizontal format, the lighting being the golden hour. Again, very detailed prompt, from the composition to goals and detailed settings.
The overall composition would have represented the very classical African savannah photos, a "real natural photograph", and the wires could have added a very strong layer to an iconic setting, but again, the e-waste is not represented the way I wanted. I was looking for a subtle layer, which this image did not provide.
Another horizontal photo prompt.
The wires are more subtle, but not exactly the way I was looking for.
Again, the composition represents The Africa, a very iconic set up, but the lighting is not strong enough. I was looking for a golden hour, emphasis on the light, rays of sunlight illuminating the exact right areas.
The lighting in this image is better, but the overall composition is further what I was looking for.
It took several reprompts, before the lighting, composition and the actual story I wanted to tell, started to form in the photo (below).
I wanted the angle to be from up, looking down.
After these, it took another ten to twenty reprompts, constantly adjusting and re-adjusting. I knew exactly what I was looking for and decided to go back towards the greenness. For this, I had several reasons, the main one being I wanted the Wasted Chances to be rich on it's imaginary storytelling as each image, but also as a whole series.
The End Result
Series: Wasted Chances
Title: Wired for Turning A New Leaf
2024
Similar type of process I go through with each photo, and depending of my little helper AI's day, it usually takes anything from ten to fifty images to get the right one. Sometimes I have to take a break, because my AI friend just really doesn't get what I am looking for, even with the numerous re-prompts.
Let me tell you, there has been some heavy re-prompting, especially with the "Behind The Scenes" series.
Page updated
Google Sites
Report abuse