The past few weeks have been filled with new studio visits with artists from around the world. This serves as both a means of identifying new artists to come into the program but also to collect raw material for a new series we are developing called Conversations With Artists About AI, abbreviated to CWA~AI!
We are still working out the exact format but the thought is to present highlights from longer form studio visits with artists in the form of Instagram reels, perhaps releasing the full discussions as YouTube videos at some point in the future. If there are any slick video editors/ social media managers out there that would like to lend a hand to this new effort, we will be standing up new social channels specifically for the AiR program and would welcome any enthusiastic collaborators who want to help engage a new Contemporary Art audience. Perhaps it could even turn it into a podcast, we will see where it leads.
The goal is to capture thoughtful dialogues with a wide array of artistic perspectives, potentially even including those who take issue with AI as a means of production. I'm want to engage openly and honestly and to have meaningful conversations with artists, curators and thinkers with vastly differing opinions about how AI currently is and may impact art production, consumption and life in general over the quickly evaporating months and years to come.
Our first conversation with Jason Bailer Losh is on the editing table right now and hopefully we will have a series of videos to release by the next update.
And on that note, please join me in welcoming a new cohort of artists!
Jason began on June 3rd and is quickly making progress towards training his own LoRAs for use in his painting and sculpture practice. Jason will be the first artist in our residency to cross the digital/physical boundary, using AI as a tool in the process of making physical paintings.
I had the pleasure of working with Jason when I ran my gallery in LA. He was part of an exhibition that I curated called As Concrete As Possible which was exhibited at Ochi Gallery, ironically in their Sun Valley, Idaho location before I moved to Idaho myself. I also was a guest on his artist-focused podcast called Seeing Is Forgetting. I've actually taken a healthy amount of inspiration from his podcast and my experience on it in thinking about CWA~AI.
Jason's work both in sculpture and more recently in painting could be described as assemblage, but for me it's more personal than that. His works extend into the realm of family history, cultural and physical nostalgia, and the role that memory plays in the creation of meaning. But, I'll let the inestimably more talented art critic Jonathan Griffin expand on his work in this snippet he wrote on Jason some years ago:
“The use of readily accessible objects and everyday materials such as foam, cardboard and plaster in Losh’s work is significant. Previous art works have involved using cardboard cast in bronze, and sculptures created from throw away materials that take on material qualities and value at odds with its humble beginnings. The detritus of everyday life is incorporated into works that at first glance appear to share and pay homage to the aesthetics of modernist sculpture. However, Losh’s practice operates across a range of registers that touch upon class, labour and craft. His works draw on biography, and the difficult subject of financial solvency. Informed by his background growing up in the small mid-western town of Denison Iowa, the economy of means by which Losh approaches making is directly influenced by an inherited biography that has been passed down to him.” - Jonathan Griffin
Jason's proposal entails the creation of a series of fungible artworks, spanning paintings and potentially sculptures. These works will distinguish themselves by their content generation, facilitated by a series of custom trained LoRAs, which will autonomously produce imagery based on the type of reference material he typically sources and edits manually. He will continuously document and share this process, utilizing both the Civitai AiR Studio and various social media outlets, serving as both a repository for source materials and a platform for conceptual sketches and studies and studio updates.
.jpg)
By integrating technology into established studio workflows, his objective is to underscore the potential for streamlining creative processes and yielding unforeseen artistic outcomes in artistic workflows that already exist. He aims to disseminate insights into this integration by sharing videos and in-progress glimpses from the studio across social media and on Figma, thereby fostering community engagement and inspiring others to explore analogous avenues of technological assimilation within their own creative endeavors.
This week we went deep on Alex's vid2vid workflow in a studio visit that was also recorded and hopefully will make its way into the CWA~AI series as well. Asad, the head of Madhatter Foundation and Basmah, the Anthropologist that we are collaborating with on this project were both on the call as well and we had an in-depth conversation about both cultural significance of the dances that are being interpreted as well as the technological workflow and how we can work to seamlessly merge these two sides of the project.
Alex has produced a few initial tests with some LoRAs that he's trained on our site. These LoRAs will have to remain private because they are trained on sensitive cultural material, and ultimately they will either be discarded or archived under the ownership of the Kalash people at the end of the project. Alex's first attempt was to replace the backgrounds of the dance videos with patterns inspired by their traditional wood carvings, completely transporting the ceremonial dance into a new visual environment. The LoRA produced a very intricate background, and I was surprised by how accurately it recreated the dancers themselves, maintaining fidelity of movements and most of their clothing. However the workflow is fully reinterpreting their dress based on the images of traditional clothing it was trained on.
Through our discussion we decided that it would be interesting to try using closeup images of their beautiful woven textiles as the background in place of the wood carvings. This was a great opportunity to test out the full workflow and see it in action, and we rendered a short clip live, which took only about 3-4 minutes including upresing. The resulting short clip is below. The workflow introduced motion into the background as well, seamlessly replicating the rippling of these stunning handwoven fabrics, along with and many of the traditional shell and bead patterns featured in their dress.
We are working with Kalashi contacts in the valley to help us record more videos and more angles of the same dance so that we can transport the viewer into multiple different scenarios perhaps including a dramatic mountain backdrop that speaks to the geography of the region.
We are excited to announce that both Ira Greenberg and Drew Nikonowicz will be starting next week!
Ira is a professor of painting, drawing, and creative coding at SMU. He is a master practitioner of generative art, having literally written books on the computer program Processing for use in an art context. Ira will also be hopping the digital/physical divide with his project, conceiving of the AI generated imagery not as an end in itself but as an augmentation of his physical drawing practice. Im excited to work with such a technologically skilled artist to produce what are potentially the lowest-tech outputs imaginable: charcoal drawings. At one point compressed charcoal was an innovation itself, and perhaps this will serve as a reminder of the historical continuum of technological innovation in art. To me, this low-tech output is a lighthearted way of decoding the present moment. In his own words:
“The image below was drawn from an automatic process, with charcoal. These images go through all sorts of iterations-some become narrative/representational and some pure abstractions/non-objective. I think of my drawings as a type of embodied generative process (originally the approach was an extension of my code-based work before we had diffusion.)
After this drawing resolved, I fed it into both Midjourney (exploring varied image weights) and Stable diffusion (varying Denoising Strength). I then referred to the AI altered images to enable me to go back into the drawing with greater insights.
So the project could be presented as an augmented approach, with a generative physical process and an algorithmic generative process informing each other, as a feedback loop. The entire body of work/process would be captured/exhibited online and IRL.”
Drew is a conceptual photographer, entrepreneur (he has his own large format camera company), prolific gamer and twitch streamer, and is currently in an MFA program at the University of Arkansas. In 2019 he released his first photo book, This World and Others Like It, co-published by Yoffy Press and FW:Books. The book explores the merging of computer modeling with analog photographic processes, drawing on 19th century survey images as an aesthetic guidepost. The resulting images - all in black and white - are intoxicatingly technical yet delicate. They somehow strike a balance between feeling highly composed yet relaxed, creating a cosmological survey of natural world and our material extension of it. The resulting nimbleness of this otherwise analytical body of work is helped by the wry humor he brings to it. Confusing in their clarity, they are actually quite fun to look at, as they require decoding to figure out exactly what is being depicted. To me, they speak to our ever increasing technological mediation of our experience of the sublime natural world.
In our initial conversation we discussed using AI as a “diagnostic” tool. I can’t describe how much I enjoy this concept (thank you, Drew!). It’s exciting to me not only because of its freshness, but because of how well it seems to describe our relationship as artists to tools and toolmaking in a integrated Studio practice, of which the emerging AI landscape is just one facet. To be diagnostic to me imparts a sense of exploration, of searching for new information, or for ways to reframe the information that we already have. The concept allows room for playfulness within a measured framework.
For example, we discussed training a LoRA on his entire iPhone camera reel over the last year, then using that to create a potential future camera reel for next year. Diffusing these “photographs” out of negative space and potentially out of future possible realities gets at the heart of that playful diagnostics he will explore during his time with us.
I love the swiftness of this idea as it speaks to the multiple layers of AI mechanics. It's literally using a predictive algorithm for its intended purpose, to predict the next token, but in this case the token is a photo that Drew might take at some point in the next year, purely based ones that he's taken last year. There's also then the ability to perpetually update the model as new real world data enters his camera, perhaps on a weekly or monthly basis. I wonder what it would take to have a perpetually training LoRA, like a living organism that is continuously evolving and refining itself. Perhaps it would even eliminate old “out of range” data on a rolling basis as it moves forward through reality, similar to how we as humans aggregate memory and experience to form our present frame of reference. I'm not sure we'll get to that point with this single project but I love the conceptual implications and the seed of possibility it represents. In Drew’s own words:
“Within my photographic practice, I work with technologies such as 3D modeling programs, 3D printing, raspberry pi powered devices, video games, and more. Technology, for better and for worse, always reshapes the world and the people around it. Then in turn we reshape technology. Around and around we go. Engaging with technology allows me to better understand myself and the world. It expands my relationship to myself and the tangible world; an otherwise deeply disorienting environment. During this residency, I will spend time experimenting with diagnostic tools specifically built around images I have personally made. My goal is to better understand how enmeshed I am with silicon based technology. Call it a digital lark.”
We are very excited to have Drew on board!
.jpg)
.jpg)
.jpg)
.jpg)