Fonema is an audio installation that explores the relationship between artificial intelligence and human communication. By processing voices through an artificial neural network, Fonema creates a distorted version of the original audio. While the voices are still recognizable, the sense of the words is lost, creating a disorienting and surreal listening experience.

The project utilizes generative adversarial networks (GANs) to generate abstract images based on military helicopter footage taken in conflict zones like Afghanistan, Iraq, and Pakistan. The use of GANs is a powerful tool, enabling the artist to create images that are simultaneously familiar yet otherworldly. The result is a collection of images that not only captivate the viewer's attention but also force them to confront the inhumane nature of modern warfare.

The installation was designed to challenge preconceived notions about how we understand and interpret spoken language. By manipulating the output of a neural network, usually used for voice classification and recognition, the work stimulates the listener's imagination and encourages them to consider the role of technology in shaping our understanding of the world. The development of Fonema involved a thorough exploration of neural network models and audio processing techniques. The distorted voices heard in the installation are reminiscent of what one might hear when not paying attention to who is speaking, questioning the relationship between technology and language.

Project for my thesis