the facts on the ground, 2021-23
The images in the facts on the ground were created in Playform. Playform boasts that it “ makes AI accessible for all creative people, without learning new terminology or learning how to code.” Using their Morph function, which takes a set of images “for inspirations and another for influences the Playform AI will create a sequence of images that transition between both sets, resulting in a mixture between your influences and inspirations.” The images I put into Inspiration were over 30 images culled from the internet of demolitions conducted by the Israeli government of Palestinian homes. The Inspiration set “informs the shapes and contours of the results.” For the Influence group I downloaded over 30 images of Jewish settlements. The Influence images will “iteratively morph to resemble this set.” Then I reversed this process so that the settlements were the Inspiration group and the demolitions were the Influence group. Images in which the demolitions are the Influence group are far more abstract. When the demolitions are the Inspiration some ominous sense of destruction remains. The bulldozers are still in evidence and the mood of the images is darker.
There is clearly a connection between the activities documented in these two sets of representations, even if demolitions and settlements aren’t in a perfect inverse ratio. It was this conceptual correlation that motivated me to create a set of images in which the two kinds of representations were visually intertwined. What does a representation of a mixture of these images mean and does any sense of their lived interdependence survive in the images that result? Can we even determine what we are looking at?
I created these images in 2021, and began to assemble them in pairs e.g. as diptychs. I would pair an image from the demolitions as Inspiration with one of the settlements as Inspiration. The meaning of the resulting image seemed to dissolve as a consequence of the visual fusion of the two related representations, one that needed too much explanation to be understood, so I set them aside. Even now, with a text, the images require clarification. What brought me back to these images was access to ChatGPT3. I wanted to create a narrative for the images to give them context but vacillated about what that narrative should be, and then ChatGPT3 became available. It was suddenly obvious that AI should create the texts as well as the images. I wrote two prompts and limited the scope of the answer to 100 words. If the response ended in mid-sentence I either deleted or cropped the last sentence. The prompts were: 1) Write a short story about two young Palestinian children playing on the beach in the Gaza Strip. and 2) Write a short story about two young Israeli children playing on the beach in Tel Aviv.
The stories are banal and quite similar with the correct assignment of Israeli vs. Palestinian sounding names. Occasionally, there is a hint of menace in the stories about the Palestinian children, but just barely. Perhaps this is more of an exercise than an artwork, given the amount of exposition the images require, so I present them with that caveat as something to think about.