Why
Echo/Everything is a place where the emergening field of artificial intelligence meets art.
The intention of the project is to create a place of blossoming and creativity where these new tools can be explored with curiosity and good intention. The discussion around anything related to creative AI is quite heated and polarised, and our hope is to take a step back from that, and focus instead on what can be good and interesting about these incredibly powerful yet poorly understood technologies. This does not exclude a critical view on any moral or practical issues, however.
We want to bring actual use cases into light, especially of the kind where the use of AI is done in close collaboration with creators involved in traditional fields of art. So far the disucssion is mostly based on fear from the side of traditional creators, and a refusal to acknowledge any problematic issues from the side of the AI enthusiasts.
Transparency is important to us, and we will try to be as open as possible about the technologies (prompts, settings, notebooks) we use to achieve whatever results will be. Currently the tools are progressing at a pace that is hard to follow, but we hope to present them so that interested artists can at least rudimentarily understand the boundaries of what is possible, and on the other hand offer detailed help for any aspiring creators who feel challenged by the messy state of affairs right now.
What
In practice what we have started to do is to train custom neural networks based on original art. We see this as a step back from the discussion over art theft, or the unsanctioned use of visual material used to train the models widely used. Of course that same training data underlies our so called fine-tuned new models too, but any resemblance to known art is buried so deep that it is not evident (unless so desired) in the eventual outputs.
We have also started to have discussions where we try to chart these new unknown lands. For a creator these technologies present a curious new experience of encountering visual material which closely resembles the original creations, but with great flexibility in interpreting it. It is better maybe to think of these simulacra as alternate universe versions of the orignal art, since at least for now the technologies are not ripe enough for rendering output which would be completely indistinguishable from the originals, depending on the complexity and finesse of any visual details and emotional charge carried by the originals.
Who
Currently the team is small, but I hope that more people can join and bring their ideas as we move towards the haze of the future.
My name is Kim, and I’m the main technical architect and initiator of the project. For a living I work as a project manager in a scitech company. I’m also a musician since the late 90s and a traditional craftsman on top of that. I’ve been interested in visually creative AIs since the original Deep Dream by Google. However, it was only when Disco Diffusion was released in late 2001 that I felt that the tools made meaningful creative work possible. Now with Stable Diffusion and Midjourney the creative possibilties are increasing faster than is easy to keep track of. I have an Instagram account where I randomly post the outcomes of my tests and explorations.
Teemu Raudaskoski, who is the first creative partner, has been interested in automated/ robotic art already before any of the current AI technologies existed, and here are his own words from his website:
“I graduated as a visual artist from the Turku Academy of Arts in 2019. My works have been showed at galleries and art events in Finland and internationally since 2002. I am currently working on painting, media art and music. The collaboration with visual artist Inka Hannula, which began in autumn 2020, has already produced dozens of co-paintings, which are on display at Gallery G12 in Helsinki and at Keskus Gallery in Tampere. Our first international show is coming to Japan in the autunum 2022.
​I am inspired by nature, especially water as constantly moving element. Painting is intuitive and physical process for me. I try to reach organic and living impressions to my works.
​I’m also musician. QOSQI is an alias for my organic electronic music. Global Hybrid Records released my Fungi album in the spring 2021.“
What about you?
We are interested to hear from anyone who would like to explore the possibilities together, and see how AI could reveal new perspectives to your already existing work. Or, you can just follow what we do, and hopefully gain some insights or just enjoy the art along the way.
We hope to let thing evolve now in a freeform and organic way, since at the pace the technology is advancing, it is very hard to have a lot of expectations about the state of affairs past a horizon of some months or so.