The idea for this article came to me after watching an interview with my favorite artist Ellen Sheidlin, who is the ideologist of a new trend in art – survirtualism (Dud, 2023). In this post, I am trying to understand, how to ‘invent’ art in collaboration with artificial intelligence (AI). I will look at the Stable Diffusion model used by Ellen Sheidlin and talk about my own experience of integration with this tool.
Stable Diffusion is an AI model that generates images based on prompts or input images and represents a breakthrough in picture generation, in terms of speed and quality (Medium, 2023). Ellen’s art is figurative and based on anatomy rather than abstraction. She used to collect collages. Nowadays, she draws an idea on a tablet and then uses Stable Diffusion to fill in the lines and explore the variations the program offers her. Once she gets a satisfactory interpretation, she refines the picture to keep her style and then transfers the result to the canvas. In this way, her craft remains personal and creative but integration with AI is added. Her vision is of co-authorship and of not necessarily abandoning the old but adapting to “the eternal addition of the new”.
My interest in Ellen’s work and motivation to experiment with Stable Diffusion software stems from a desire to explore the intersection of AI and artists. Therefore, to expand my understanding and creative boundaries, I decided to experiment with the web interface of several web pages offering access to the Stable Diffusion model. I used the ‘Stable Diffusion Online’ and ‘Hugging Face’ platforms, which offer free access to image generation.
Since Christmas is around the corner, I decided to write a prompt about Christmas markets in the Netherlands. The initial request gave quick results, however, I decided to complicate my prompt to get a visualization I couldn’t find on the internet. I added that everything should be in pink colors and asked for a Christmas market on the beach. The results were impressive. I see the potential to use this AI system for my studying and everyday needs. However, I agree with Ellen that for artists it serves more as a mockup. The results obtained on the ‘Stable Diffusion Online’ platform (pictures 1 – 3) were better than ‘Hugging Face’ (pictures 4 – 6) in terms of quality and playback speed. The Stable Diffusion model is perfect for specific requests, bringing to life the most unique fantasies.
In conclusion, I strongly believe that AI is a powerful tool for inspiration. Art and AI may compete, but in this competition, a new market is born. I think that AI can be a tool that can help unlock creative potential. I would encourage everyone to explore Ellen’s work and experiment with Stable Diffusion to realize the full potential of this creative work.
References
Dud, Y. (2023, September 06). Ellen Sheidlin – global art star from Russia. Retrieved from YouTube: https://www.youtube.com/watch?v=3TEKiX9UoY4
Medium. (2023, September 06). Understanding Stable Diffusion: The Magic Behind AI Image Generation. Retrieved from https://medium.com/@amanatulla1606/understanding-stable-diffusion-the-magic-behind-ai-image-generation-e834e8d92326
An often discussed aspect of generative AI images is that it has the potential to replace art made by actual humans. However, in this post it is clearly highlighted that it does not have to be one or the other but that both these mediums can be used together. I have personally never heard of Ellen Sheidlin and it was new to me to see this mix method of using both AI and people to create art together. It was nice to see that you explored with Stable diffusion yourself and shared your personal experience. Thank you for exploring more into the topic about human collaboration with artificial intelligence to create art.
It’s interesting to see how artists can use AI as part of their artistic process and as inspiration. Based on your recommendation, I’ve looked up Ellen’s work online, and I was definitely impressed on what AI can do! I did see some conflicting comments on her Instagram from other artists who believe AI is an exploitation of “real” art, and that artists shouldn’t promote AI like this. Do you think AI will benefit or harm artists in the future? For example, the AI photos you’ve created can easily be transcreated and used by companies for promotion around Christmas, meaning they wouldn’t need to hire a photographer anymore for campaigns. I’m interested to hear your thoughts on this.