︎
G H O S T S
March, 2020
Today, images of faces that appear real but do not actually exist circulate widely. This is made possible by GAN algorithms (Generative Adversarial Networks), artificial intelligence systems capable of recognizing facial patterns and generating realistic portraits of non-existent people. However, when the system fails, the faces collapse into visual chaos.
Ghosts is a project that combines the traditional technique of oil painting with the computational possibilities of GAN networks, creating images born from algorithmic errors: artificial, unreal portraits that evoke spectral presences.
These beings — made of bits and paint — emerge from a dialogue between the machine and the artist. Ghosts offers a retrospective gaze at technological evolution, reminding us that in the visual universe, there are always paths that lead both forward and back.
Today, images of faces that appear real but do not actually exist circulate widely. This is made possible by GAN algorithms (Generative Adversarial Networks), artificial intelligence systems capable of recognizing facial patterns and generating realistic portraits of non-existent people. However, when the system fails, the faces collapse into visual chaos.
Ghosts is a project that combines the traditional technique of oil painting with the computational possibilities of GAN networks, creating images born from algorithmic errors: artificial, unreal portraits that evoke spectral presences.
These beings — made of bits and paint — emerge from a dialogue between the machine and the artist. Ghosts offers a retrospective gaze at technological evolution, reminding us that in the visual universe, there are always paths that lead both forward and back.
GHOSTS
Fixed and moving images produced with artificial intelligence.