Bugs in the system

My research focused on exploring AI’s understanding of knitting and knitted fabrics. I used various image generation programmes, issuing simple commands like “knitted fabric” or “knitted material”. AI produced images that depicted knitted textiles realistically; however, when asked to do further variations on a given image, the AI became confused and made errors, generating unrealistic images. While in real life, errors in knitted fabric typically involve tears, dropped stitches, or holes, thus representing some form of deficiency, the AI interpreted errors as congestion or surplus. Using this as a starting point, I started deliberately generating flawed images that served as inspiration for my experiments with knitted fabric, highlighting surplus and congestion. Building on these material experiments, I designed a 12-piece collection, deliberately incorporating the errors made by the AI not only in the materials but also in the silhouettes. Ultimately, two outfits from the planned collection were completed.