Algorithmic Aura

This work explores how machine learning processes can evoke moments of uncertainty, emergence, and ecological memory. These images were created using generative adversarial networks, or GANs, trained on endangered insects and flowers in Ireland. Held back before reaching photorealistic fidelity, these images preserve evidence of loss and transformation, intentionally halting the machine training process to maintain an aura of incompletion. The resulting images, like butterflies still halfway emerged from their chrysalises, invite the viewer to imagine the next stage without ever resolving it.