Abstract

Creating a mobile software-based exploration (artwork?/app?) puts the artist-coder in a position to interact with the mediated image streams that connect people on the Internet. The mediated streams often contain portraits and self-portraits, selfies, of the participants. These selfies are visual status messages of the people participating in the data streams. They can be used by the poster to identify themselves in the data stream and represent a way the creator of the selfie wants to be seen by the social circle in which the image is posted. These images could reveal recognizable characteristics of the subject. The characteristics could be recognizable by people and software along the mediated path that the image traverses. Algorithmic intervention on the image affects the human and software interpretation of the image. In the works #autoselfie and FIFO ribbon, images are modified on a pixel level, in a way that the user of the app/work cannot control completely. The modifications reveals the potential power of the software systems to modify the imagery that people use to identify themselves in mediated communication environments. Moreover, the changes applied are chosen by the artist coder to impede recognition by software systems downstream but make human recognition possible in some cases. The alterations make the software in the app visible on a pixel level. The works then let the user post the altered images on Twitter or Google+ social media streams. The framing of this work as artwork or mobile app creates a tension, which affects audience interpretation and user experience choices made by the artist. These ideas are explored in the 2013 work #autoselfie and the 2014 work FIFO ribbon that runs on Android devices. #autoselfie was used to remix images created by winners of the South African 2013 Sanlam National Portrait Award, to explore the intervention of algorithms on mediated portraits and self-portraits.