This talk will introduces ITERATIVE BODY SYNTHESIS, a media installation that acts as both a research infrastructure for an experimental online investigation and a neural backend and monitoring interface for a virtual identity on Instagram. The project aims to investigate the impact of algorithmic decision-making on body representations in social media, exploring how these platforms may shape and distort forms of representation and perception such as body aesthetics, choreography, iconography, and authenticity.
Social media platforms like Instagram treat their users as black boxes, analysing their affects based on continuously sampled visual media. By reversing this practice and identifying algorithmic “affects” towards specific body characteristics, we cast a virtual hyper-embodiment of Instagram’s affective economies and notions of integrity. To this end, we have developed a prototype of a self-operating software architecture based on principles of black box testing, machine learning, and synthetic data. With this, ITERATIVE BODY SYNTHESIS explores technological possibilities for developing a digital infrastructure that empowers citizens to collectively monitor algorithmic systems that moderate, filter, and verify our media realities.
This work in progress presentation will move through the different trials, challenges and (preliminary) results of the project, elaborate on areas for further research and provide an overview of the project’s ongoing and future developments.