Grebenshyo グレベンシ

@grebenshi ∙ Published February 16, 2024

What’s your background and how did you get into AI art?

I have a background in traditional art, having graduated in painting, but even back then I was specializing in mixed media, using oil paint alongside photography and digital tools. Since then, I have completely moved to computer-based production, gradually shifting through generative art, digital painting, and 3D modeling.

I’ve been interested in the concept of “computationality” for many years, but since I’m mostly developing a personal project that is not centered around techniques, I never really understood what the peculiar images filled with dog eyes (I’m looking at you, DeepDream) were all about, until the generative approach of diffusion became mainstream a couple of years ago.

One of the major inspirations at that time, though not the only one, were the fantastically intricate pieces by Noizebloc, whom I still consider one of the realest masters in the field to this day. Being accustomed to this route through my experience with Houdini and similar software, I quickly turned to YouTube tutorials, and not long after, I was paying for my first Google Colab subscription to use a DiscoDiffusion notebook.

What drives you to create?

To me, art is the vehicle to reflect the surrounding world. A way to represent what I see around me in the manner that it happens.

What does your workflow look like?

As I previously mentioned, I’m proficient in various techniques, which have enabled me to train my own AI models using 3D renders I’ve created, among other things. In fact, I view AI more as an addition to my toolbox rather than a replacement for other methods.

I’m also an avid user of the ComfyUI node system, which facilitates the creation of custom workflows. For example, I recently locally replicated the EmProps OpenStudio’s AI to p5.js pipeline for a project I developed on the platform, in an effort to increase the resources.

Another crucial part of my process involves complex in- and outpainting tasks that include digital painting and related activities: image editing, collage, post-processing, and so on. A vital concept in AI art, known to many serious ‘pros’ in the field, is the so-called autopoietic approach. This involves feeding the system its own outputs to refine the results through several steps, approximating a form of “control” with regard to this specific tool.

How do you imagine AI (art) will be impacting society in the near future?

In the near future, the impact of AI is likely to be as significant as the advent of the internet, letterpress printing, or the Industrial Revolution. It will be an inescapable force, profoundly and radically altering every aspect of our activities…

Looking a bit further, yet still within a relatively “near” future, I believe AI and AI art will be instrumental in transforming the very fabric of our society. This transformation may trigger processes akin to those seen during the Enlightenment or the Renaissance, reshaping everything from politics and philosophy to science and beyond. The world as we know it today will be fundamentally changed, in both visible and subtle ways.

Who is your favourite artist?

Leonardo, Michelangelo, Donatello and ~Rafaello~ Albert Oehlen ;)

What is your favourite prompt when creating art?

I suspect that the culture of “prompting” might be heading toward obsolescence. This is becoming apparent due to the semiotic discrepancy noted in the “image-to-prompt” process in model conditioning. Simply put, using images as inputs for image synthesis, instead of words, might yield “better” or at least “different” results. This shift amplifies the focus on the differences between these methods, rather than on the peculiarities of each individual one. However, I still actively use so-called “trigger words” with custom models. These are usually “nonsense” words that can’t be found in vocabularies, in order to not conflict with training data, such as “zwx” or “sks”.

Do you have a specific project you’re currently working on?

Most of the time there are a few parallel ideas/projects I’m developing, and it’s not an exception currently. The one I’m considering being the most “important” one however, is a new installment on EmProps that would involve a pretty extensive use of the img-2-img method. I can’t reveal much more right now though :)

Would you tell us about the AI Surrealism exhibition in NYC known to be the largest AI art showcase in the world up to date?

Well, first of all, AI Surrealism exemplifies the new opportunities AI art presents. While it’s a thematic drop and show, it has also been a consistent and high-quality release drop, both from the standpoint of content as from that of the form. AI art, being what it is, facilitates communication in the field, fostering cooperation and discussion around relevant cultural topics, taking the form of various original initiatives and events.

With 100 artists involved, AI Surrealism has brought this concept to a new scale, one of the exhibition’s main appeals. Regarding my contribution, I view Surrealism as the most recent significant aesthetic statement in art, likely revitalized by what AI can offer. My pieces aim to pay homage to the high technical standards that characterized the movement.

What does it mean to be an AI Surrealist for you in the times we live in?

During the Twitter space dedicated to the show, I mentioned how, in my opinion, Surrealism and AI art are united by the idea of collectivity. This is reflected in the relationship between the notion of “collective unconscious” in Surrealism and, from a data analytical perspective, in the context of AI. Relating this idea to the surrounding world is what I consider the essence of “AI Surrealism.” While it’s no secret that movements in art have long since faded, initiatives like these remind us that true art aims to transcend time, bringing the noblest ideas to life.

Would you like to share anything else?

Many thanks to everyone who contributed to bringing this fantastic event to life, from the curators of Anna Dart and Roger Haus, founders at the Exquisite Workers Collective, to the Superchief Gallery NFT, to the platforms hosting, covering, and promoting it and of course all of the amazing artists participating.