In this project, I explored an idea of co-creation by an Artist (me) and an Artificial Intelligence (AI) algorithm. If you maybe have heard, AI shows good results in cancer detection, face recognition, analytics, and so on. But, you should have heard that the best results we get when we combine AI with real human experience. I decided to play around a bit and see what we would get together.
We made our moves turn by turn. I used a photo I took and fed it to AI. The result was not very different from the original at first. The algorithm saw the landscape and reproduced the landscape accordingly. Then I tried to add to the image something not clear to the algorithm. I drew some new objects in the photo and gave it back to the algorithm. Back I got a photo similar to what I gave the algorithm from the beginning. It seemed that the AI simply removed everything he did not understand and brought the image to the standard. But I kept adding more and more elements to the photo.
Eventually, the algorithm began to generate more unusual images. These images turned on my imagination. In obscure patterns, I saw objects that my brain recognized. I repeated this action in 10-15 steps and now I publish the images we got with the algorithm during the game.
In the process of creating this project, I often realized how similarly the algorithm and the person recognize the image. Each of us subconsciously recognized the objects he understood in a set of abstract spots. The only difference was in experience. The algorithm was trained on thousands of images, and my brain was trained on an incredible number of objects seen in 30+ years of my life.