Info     Blog     E-shop︎ 

︎    ︎    ︎


artistic research, custom StyleGAN model, generated visuals, 2021

TroublingGAN is a critical generative artwork and artistic research project that experiments with the StyleGAN vector-image generative neural network.

TroublingGAN is a StyleGAN︎︎︎ model that generates visually ambiguous visuals representing "troubling times". With the goal to create a different kind of knowledge about this elusive concept, we used the generative neural network to process uncategorised news photography from 2020, define specific visual features and generate new examples of recognised patterns from inside the dataset. TroublingGAN is able to derive an essence of "troubling times" and project the affective value of news photography (as used in our training dataset) onto generated visual outputs.

TroublingGAN should not be understood as a final outcome. It was created to be used as a tool for artistic research to create an alternative perspective on how to view the images that were used in the training dataset, pointing out the more subtle knowledge previously hidden behind the concrete details, context and meaning of each image.

scrolling through the dataset of 2020 news photography

The observed object - the concept of “troubling times” - is inspired by ‘Designing in Troubling Times’, the theme of the 2021 Uroboros Festival︎︎︎. As “troubling times” we understand rapid socio-ecological changes caused by shifts in global economic, political and technological power and the subsequent series of troubling events, including the COVID-19 pandemic, violent conflicts and environmental catastrophes. Early outcomes from the first StyleGAN training were incorporated into the visuals of the 2nd edition of the festival:

Although TroublingGAN images hardly consist of specific objects or symbols, they strangely resemble photography and, however abstract they may be, the human mind wants to impose meaning on them. Instead of what they represent, we focus more on the affective quality of these synthetic visuals and speculate on their potential use as a replacement for photojournalism. Images documenting tragedies and catastrophes are so over-used in contemporary culture that visually overloaded viewers become indifferent to their content. We suggest replacing photojournalism with visually ambivalent synthetic footage in order to achieve greater involvement with the attached text message.

snapshot from the training

This artistic research is a critique of the techno-solutionism and mystification of Artificial Intelligence, focusing on generative neural networks and questioning the value that this technological advancement can bring, outside of commercial applications.

This project, on one hand, highlights the use of GANs as a research tool - a tool of visual analysis The generated visuals, although visually ambiguous, have a potential to create an alternative perspective on how to view the images from the training dataset, pointing out the more subtle knowledge previously hidden behind the concrete details, context and meaning of each image.
On the other hand, through the provocative use of TroublingGAN outputs instead of news photography, we point out the unexpected relevance of ‘imperfect’ outcomes in AI-driven visual synthesis. Visual synthetic media, whether photo-realistic or visually ambiguous, forms a completely new category of visual material, and its place in visual culture has yet to be determined. Its spectacularity seems to be a temporary effect caused by its novelty; however, the anxiety of its indefiniteness and its affective quality are features of its AI-generated origin and need to be accounted for when working with these visuals.


Concept and research: Lenka Hamosova

Technical guidance: Pavol Rusnak

Scraping magician: Adriana Homolova

Original code: StyleGAN2 (Karras et al.)

FAMU Prague, 2021

This work was created at the Academy of Performing Arts in Prague as part of the project "Extending the creative tools of Machine Learning and Artificial Intelligence - Experimental Tools in Artistic Practice" supported by the Ministry of Education and Science for specific university research at the Academy of Performing Arts in Prague in 2021.