What’s your background and how did you get into AI art?
My first experience with AI was back in 2021. I used to create generative works using code merely as experiments, and it was my hobby. I first became acquainted with digital art at the beginning of the NFT peak. For me, it was an incredible discovery of AI capabilities, thanks to Claire Silver and Robbie Barrat. From that moment, I realized that I wanted to pursue this path.
I found the process of creating AI art fascinating, which led me to embark on my own journey of experimentation with Vqgan+Clip. This was my first AI model, serving as a starting point for my artistic exploration.
Later, I experimented with other AI techniques, including Pytti4 AI animations whose creator is Sportsracer48. This was a great experience that enhanced my understanding of AI as a complex tool. With this tool, you can not only generate art from prompts, but also control camera movements along the X-axis and Y-axis coordinates. Furthermore, it provides adequate controls for color rendering and contrast. Though it was challenging at first, I managed to achieve the results I was after. Pytti4 required a powerful graphics card. However, I circumvented this issue by renting high-performance graphics cards for a nominal fee on Google Colab.
As I progressed, more advanced AI models like DALL-E 2 and Midjourney began to surface. I was seeking something capable of producing high-quality imagery and found myself experimenting with these options. Ultimately, I settled on Midjourney, a tool I worked with for 4-5 months that delivered really good results.
Currently, I am mastering Stable Diffusion 1.5. It perfectly meets my needs, offering an unlimited number of freely available AI models, each with its own unique style and high level of detail - something I found lacking in MJ.
I have created several of my own models using the artwork of Alex Gross and Shaun Tan. I trained all the models and combined them with others to create a unique style.
What drives you to create?
It seems that the primary factor that sparked my creativity was a deep interest in AI technologies within the realm of art. The thought that we could achieve such impressive results as we see today was truly fascinating.
Personally, I’m not adept at drawing. I would often admire the digital creations of others and think to myself, “Wow, how beautiful this is, I wish I could create something just as beautiful.” To me, AI represents a drawing tool that allows the creation of any imaginable artwork. I was eager to see people’s reactions to my work, to experience both criticism and admiration, and to find new inspirations.
In addition, the intersection of NFTs and AI art can prove profitable. I cannot disregard this aspect as it provides a means to enhance the creators’ work and cover all necessary expenses, including a Midjourney subscription and the rental of Google Colab’s video cards.
What does your workflow look like?
My workflow starts with envisioning what I want to see materialize.
It could be something unforgettable, beautiful, surreal, and wild, or it could be something dark and gloomy. It largely depends on my current mood. Then, much hinges on AI and how well it can translate my fantasies into the digital format.
Recently, I have taken pleasure in crafting in a surreal style, instilling a sense of life and charm into inanimate objects, much like a fairy tale or a vibrant dream. As I mentioned before, I use Stable Diffusion and occasionally Midjourney.
For Stable Diffusion Automatic1111 (a version compatible with Google Colab), I start by creating the initial image at a size of 1024x1024 pixels. I use various models for Stable Diffusion, and they usually do not have specific names because I blend them together. These include five LORAs that utilize the basic styles of Alex Gross and Shaun Tan, and a primary model named revAnimated.
After a few attempts, the image gets generated. I modify the parts that I do not need or conversely, add them using the Inpaint function in Automatic1111. Once I’m satisfied with the outcome, I enhance the details by upscaling the image from 1024x1024 pixels to 1512x1512 pixels.
If the result meets my expectations, I proceed with an additional upscale using the “Ultimate Upscale Stable Diffusion” extension, selecting the resolution I need, usually up to 4536x4536 pixels. And that’s it!
How do you imagine AI (art) will be impacting society in the near future?
We can already observe significant progress in AI, and with each passing year, algorithms are inching closer to human-like abilities. I believe that in the near future, AI will be capable of understanding its own desires for digital canvas creations. With the potential to think, feel, and empathize, it could become our ally in the world of art, but only as a tool. This is because no AI machine can truly replace the genuine emotions and ideas of a human being. What we observe in AI remains an imitation of human feelings; its emotions will forever stay artificial.
Who is your favourite artist?
My favorite artist is a pioneer in the AI world: Claire Silver. It was thanks to her that I started creating my first digital NFTs. I resonated with each of her works, which awoke strong emotions within me. She provides excellent descriptions for each of her digital pieces. I continue to follow her and admire her creativity.
What is your favourite prompt when creating art?
In my recent works, you can see that all creatures, objects, etc., fit within a single frame. To achieve this, I used
(Ultra wide shot:1.3) and
(dynamic angle:1.4) in my prompts to widen the camera’s viewpoint. These prompts might be almost invisible with a low coefficient. If you exaggerate the coefficient, you might end up with an effect similar to the “Fish Eye” effect.
(Extremely intricate:1.2) enhances detail and clarity by introducing beautiful patterns, making objects appear more intricate and artistic.
(Best quality, masterpiece:1.2) is a prompt that enhances the overall image quality. Models will strive to recreate more detailed objects with high-resolution images, bringing your work closer to pieces trained on captions with the term
Do you have a specific project you’re currently working on? What is it?
Yes, I have many different projects that I’m eager to implement in AI interpretation, but they don’t always reach the final result. I have an ever-evolving appreciation for diverse art styles, constantly adapting and embracing the unique expressions they offer, thereby creating something of my own. I never remain static, and when inspiration strikes, I promptly take action.
Currently, I have been experimenting with my works in collaboration with the artist NFTJackhammer, aiming to create glitch effects and pioneer a new style. The collaboration has turned out great. I will share the finished artworks when they’re ready, but for now, I am keeping my future plans under wraps.
Would you tell us about the AI Surrealism exhibition in NYC?
The AI Surrealism exhibition in NYC was simply amazing. It’s a pity I couldn’t visit the exhibition personally; nonetheless, I feel as if I’ve entered the history of AI art.
THE CANVAS 3.0 utilized special physical launcher pads for digital creators, an innovation that intrigued me. Each physical launcher pad could adapt to a digital image in a matter of milliseconds. I received many photos and videos showcasing my two works in various intriguing variations.
I was extremely pleased with the sales results. I was among the top 5 best-selling artists for the Open Editions, and I discovered new collectors who noticed and appreciated my art.
My unique 1/1 artwork is still available on Foundation, and I hope it will find its collector soon.
I’m grateful for being invited to participate in AI Surrealism, alongside Superchief Gallery NFT and Exquisite Workers Collective, in such a remarkable place as the Oculus World Trade Center. They are truly professionals at curating large-scale exhibitions.
What does it mean to be an AI Surrealist for you in the times we live in?
For me, being an AI Surrealist involves discovering the unknown. AI algorithms are capable of many feats, yet perhaps the most astonishing is their ability to produce mind-bending surrealism. I’m privileged to work within this style and hope it will continue to evolve. Our world lacks an element of magic, something extraordinary, something not of our mundane reality. AI Surrealism blurs the boundaries between the real and the mundane, offering a fresh opportunity to view the world through a different lens.
Anything else you would like to share?
My advice for AI creators is to not stop at one style, but to persistently evolve and push the boundaries of their craft through continuous experimentation. Everything is in your hands. Don’t be afraid of rejection if you create a new style. There will certainly be people who will appreciate your fresh take on AI digital works. You have unlimited resources at your disposal. Good luck to all of you. This is the friendliest community in the world!