Tezos

Obfuscated Ascent of Ilium’s Fabricated Mind

Creator

tz1eZ8amacCvSQsFM8wampwWXJFWsmGVRQFd

Description

In the early days of my practice with emerging algorithms in 2019, my primary objective was to achieve a photographic level of quality in my outputs, as this represented the most challenging frontier in digital art at the time. This pursuit of the most difficult path has always defined my approach to art, contrasting sharply with my earlier struggles to find discipline in traditional studies and creative expression. The exploration of synthetic image generation through deep-learning technology, often mistakenly referred to as artificial intelligence (AI), has provided me with a medium to reconcile these conflicting aspects of my nature. This technology has become the perfect tool for navigating my innate challenges with structured discipline and conventional methods, which have historically held back my artistic expression. Through AI, I have found an empowering way to overcome these limitations, achieving a harmonious blend of meticulous detail and creative spontaneity. As generative AI evolves rapidly, my focus has shifted from pure realism to exploring the edges of reality—those ambiguous spaces that lie between the physical and the imaginary world. My work now aims to evoke surprise and wonder, inviting the viewer to complete the artistic journey. It is in these pseudo-figurative pieces where I hope my work transcends its medium and finds its true form within the viewer's mind. As a professional photographer, I have always understood that true artistry goes beyond the mere capabilities of technology. It is essential that my work reflects my personal vision and character. While painters and sculptors naturally infuse their work with their essence, the challenge is greater when the method of production, such as a camera or an algorithm, plays a significant role in the final outcome. In these contexts, maintaining a distinct personal signature in one's work is crucial, ensuring that the art reflects the unique perspective of the artist rather than just the tool or technology used. Since early AI models did not utilize Large Language Models, the practice now commonly known as "prompting" was unavailable when I began working with these tools. Generating outputs primarily involved adjusting weights and mapping an uploaded image into the GAN’s creative space, enabling the AI to recognize and create variations of that image. I’ve continued to use this approach, not only to integrate my photography into new creations and keeping tighter authority over the outputs, but also because depending solely on simple prompts feels like relinquishing too much control to the machine, which could make my work easily replicable by others. My methodology also involves constantly reworking generated images using different tools and algorithms, which adds a rich and layered palpable history to each piece, and ensures that the results feel original and not produced by any single, recognizable method. | artist: Bas Uterwijk (@ganbrood)| PNG, 2691x3840 pixels | original creation date: May 2, 2024