Chloe McFadden
🌟 Work
🌟 Contact
Generated Tools (2021-2024)
Generated Tools is a practice-based research project in which I worked with a Generative Adversarial Network (GAN) to create an installation that features tools as subject matter. In this project I was exploring how artists can make strange with machine learning in ways that allow audiences to critically reapproach the known and familiar. Via the creation of strange tools, the differences in how people and algorithms ‘conceptualise’ usefulness is engaged and probed.
The installation consists of a tool wall which houses three instances of working with a StyleGAN2 trained using a dataset of images of hand tools sourced from the Bunnings Warehouse website (an iconic Australian hardware store). The first instance shows 3D printed tools, modelled and printed using the generated images by the GAN as reference. The second instance shows 2D “latent space walk” videos projected on the tool wall, created by incrementally sampling the latent vector space generated by the GAN. The final instance shows doctored Bunnings Warehouse product catalogues, featuring images of tools generated by the GAN. Audience members are encouraged to engage with the work by flipping through the product catalogues, as well as being able to pick up, play with and rearrange the 3D printed objects on the tool wall.
Machine learning applications that can generate images is compelling technology - sparking new ways of making and creative possibilities. However, it also has the potential to be a reaffirming conservative force, as it is informed by the training dataset and may familiarize and perpetuate the visuals, ideas and assumptions hosted within. While this reliance upon the dataset may pose a risk if we engage uncritically—this also presents the opportunity for artists to disrupt, explore and denaturalise the known.