AI UNIVERSE

My sculptures’ journey into the digital realm, in Collaboration with Nodenaysteen

The exhibition runned from the 27th of October till 17th of December, 2023

The "AI UNIVERSE" exhibition brought my entire sculpting practice to life in the digital realm of AI. This exhibition was supported by Nodenaysteen.

As a traditional sculptor who primarily works with old-school handmade sculptures, I ventured into the world of offline-generated AI, specifically exploring the realm of Stable Diffusion. This exhibition showcases the possibilities of crafting your own universe through the full power of personalized AI at your fingertips.

"AI UNIVERSE" begins with the past creation of clay sculptures, and these serve as the foundation for my AI models, which I personally trained. AI became a co-creator, generating images that resonated with my artistic vision. For example, my "satellite sculptures" underwent astonishing transformations under the influence of AI. They transcended the bounds of reality, floating through space, exploding into vibrant displays, and moving faster than light—all without an online connection.

The process in "AI UNIVERSE" harmoniously blends handwork with offline AI technology. After the AI generated images, I stepped in with digital hand drawing tools to further refine the results of the AI.

"AI UNIVERSE" is a journey where my sculptures give birth to captivating digital creations, where my hand meets AI, and where art finds innovation. This exhibition challenges conventions and redefines what's possible in contemporary art, all while highlighting the special nature of offline AI.

See you on the opening night on the 27th of October! The exhibition will run until the 3th of December

AI UNIVERSE
In cooperation with Nodenaysteen, Gallery for Art and Technology. You can find more on Nodenaysteen’s website: www.nodenaysteen.org

27th of October - 17th of December, 2023
On Friday, Saturday and Sunday, or on appointment (via contact page)
From 2pm till 6pm
Nodenaysteen, Predikherenlei 4a, 9000 Gent, BE

The development history and remarks of my AI discoveries.

In February of this year (2023), I discovered ChatGPT, much like many of us did. What immediately struck me was its immense power. In just one day, collectively, we uncovered a force that was unprecedented. It quickly became apparent to me that I needed to explore and delve into AI myself. I feared that without a sufficient understanding of what it is and how it works, I would face the same technological gap that my parents experienced with computers and the internet—a gap that seemed insurmountable. That couldn't happen. AI isn't a matter of good or bad; its power compels you to comprehend it simply because the rest of humanity is doesn’t have a choice. It’s just that powerful. That was my motivation to investigate AI.

A few days later in this research journey, I encountered Stable Diffusion. Similar to Midjourney, Stable Diffusion is a model that enables the generation of images. For example, if you want an apple, the model creates it. If you desire a golden apple, the model crafts it. If you seek a man of 1m75 with a brown fedora hat, black Oxford shoes, and a crease in his trousers, the model produces it. Let that man float through the cosmos, and the model generates that. Ask for a sculpture of myself in the cosmos, and the model does not create it because it is unfamiliar with that specific sculpture.

Stable Diffusion operates in a mathematical manner (using vectors, not images) where archetypes are embedded. The model is pre-trained on diverse subjects from our world, incorporating mathematical descriptions of apples, people, and even styles such as those of Vincent Van Gogh or Rembrandt. This allows you to create fictional scenarios in the styles of different artists.

From the moment I started using Stable Diffusion, it became clear that I needed to train such a model myself. Training is jargon of the process of incorporating new objects and images into the model. My sculptures needed to be included. My ultimate goal was to create a model encompassing my entire sculptural practice, enabling all my creations to transcend the limitations of our tangible world and come to life in the realm of AI. Initially, I focused on training one theme/series of sculptures for the "AI Universe" exhibition, specifically my latest series portraying satellites.

Satellites serve as an excellent metaphor for what I'm striving to achieve in AI Universe: they represent our initial step towards the exploration of the cosmos. By training satellites, you can precisely depict them and even assign a class, such as "satellite." This opens the door to new sculptures following the principles of my visual language, without physically creating them, thus stimulating my imagination.

Stable Diffusion is an offline AI, meaning it is not connected to the internet. Unlike models like Midjourney or ChatGPT, which operate on servers of large companies, this AI works on your own computer. This data is not shared. Google, Facebook, and no one else can access it. This model is solely trained on your own input, making it an integral part of yourself.

To train Stable Diffusion, I purchased the most powerful computer in mid-April 2023. My aim was to organize an exhibition in November 2023 based on my discoveries: "AI Universe" (In collaboration with Nodenaysteen). As I delved into it, I noticed that AI had an intrinsic influence on my imagination. The model enabled me to visualize things in my mind that would never have occurred without the use of artificial intelligence.

The exhibition showcases sculptures trained on AI alongside a selection of AI output that I continued to work on. The primary goal was not just to exhibit pure output but to present finished artworks where the spectator experiences a coherent visual language, even with AI involvement. Contrary to my expectations, I quickly succeeded in achieving this.

In the early stages of my discovery process, my output was limited by my own imagination, resulting in sculptures that quite literally ventured into the cosmos. After all, I am a sci-fi futurist. As I progressed in the process, my output became more autonomous, surreal, and dystopian but still within the same energy field of my existing artworks.

For the exhibition, I generated a batch output of images using AI. Subsequently, I selected images and applied color adjustments. The shape and composition were primarily determined by AI, after which I made manual adjustments. After the color adjustment, I sent the image to a drawing program (Krita, similar to Illustrator), where I continued to draw manually on a separate drawing layer. This process, using a digital drawing tablet suitable for ultra-high-resolution images, allowed me to make artistic decisions regarding form and texture. The entire texture visible was of my own creation. In other words, the AI process was sandwiched between my handmade process.

Initially, it was thought to be difficult, if not impossible, to develop a unique visual language with AI. However, I learned that substantial efforts yield substantial results. Ask complex questions, and you receive complex answers. Infuse something of yourself, and something of yourself emerges. For an artist with a substantial practice, this is a boon. In my case, the output of AI cannot exist without the actual sculptures I've created.

AI opens up everything for me, leads to new discoveries, is an integral part of my handmade creative process, and fosters a creative interaction. The faster this interaction occurs between myself and AI, the better and more natural the results.

The goal is to broaden my imagination, not just to create output on 2D mediums like prints and screens (as in the AI Universe exhibition) but also to return to my handmade sculptures. In the future, I envision shaping my output based on AI trained on my sculptures.

It has become clear to me that the same principles apply in the creative process with AI as in the creation of visual art. It involves adding, subtracting, and tuning into yourself. The intriguing aspect of AI is that we don't actually understand how it works internally, but by working extensively with it, you can sharpen your intuition. The intuitive process of each artwork is also present in the methodology of AI: the "why" is not explained, but it becomes an innate process where you sense the AI and accurately determine the direction.

In conclusion, an interesting example: for this exhibition, I focused on satellites. Until then, my sculptures had straight solar panels. However, thanks to the output of AI, I saw triangles and tilted pieces being incorporated into the solar panels. This idea was fantastic and phenomenal. A good idea is a good idea, so I used it. In the AI Universe exhibition, you will see my first physical sculptures with solar panels, inspired by the AI output. This illustrates how AI can influence who you are and what you do.