Assisted Self-Portrait of an AI

8 works from the series Assisted Self-Portrait of an AI

The pieces belonging to Assisted Self-Portrait of an AI collection were produced by artificial intelligence, utilising the text-to-image machine learning technology known as Stable Diffusion, specifically with the Stable Diffusion XL 1.0 model. The images were generated around the following prompt:

Portrait against a white background

Since the advent of widespread generative AI, I have been intrigued by the concept of an AI’s portrait. Could we create a portrait that goes beyond the clichés seen in popular media, revealing an aspect of the AI’s character? And could this be achieved using the AI’s core generation process, free from human biases?

This fascination led to the creation of Assisted Self-Portrait of an AI, a series of 8 photographic AI works. The work explores the idea of an AI creating its own portrait, considering what form this self-representation might take and the mediums it might use.

Using text-to-image generation with the prompt “portrait against a white background,” the images were generated in a single iteration with one denoising step, utilizing the text as both positive and negative input. This approach forces the AI to reconcile conflicting instructions in a single creative act. The generated images were then finalized using traditional photography processing tools.

The collection comprises a total of 8 pieces.

Specifications

Nr of works: 8
Black and White / 1:1 ratio