In the Data Games series of projects, the author investigates the speculative nature of computer algorithms that create new potential forms, types and spaces with the help of generative adversarial networks (GAN) and through the synthesis of images from existing publicly available collections of 3D models. For the purpose of a series of projects, an application was developed that can learn and create new 3D models using machine learning and computer data processing technology. In a multi-year process, the author intends to explore a number of possible conceptual frameworks that will share a common thread and question the use of computer-generated data in the creation of possible alternative historical narratives.
The first conceptual framework, called Stones, used an extensive dataset of scanned historical artifacts, specifically busts. This collection of 3D scans of busts was obtained by the author in cooperation with the Scan The World platform, which takes care of documenting and storing historical artifacts of the most important museums and galleries around the world. The bust is an art form that has historically dominated the presentation of eminent figures and dignitaries from a wide variety of fields. The results of the generated models, which are presented in the fictitious meta world, reflect poor gender, cultural and racial representation, which stems from already fundamentally historically biased museological practices and deficient collections in modern museums.

Author: Anže Sekelj
Programming and Development: Benjamin Fele
3D design and animation: Luka Grčar
Collection of 3D models: Scan the World
Co-production RUK/DDT