Prepare for a fresh edition of Stable Digest that won't disappoint.
Catch our StableChat AMA on YouTube and seize your final opportunity to snag CogX passes!
This week's spotlight features the incredible KEZIAI as they navigate both chronic illness and the creative process.
In Tools & Tuts, discover the ins and outs of Gaussian Splatting, along with highlights on community projects 4MOOF and Reverie.
To cap it off, we're celebrating the awe-inspiring contributions from our gifted community. This is one issue you'll definitely want to read!
Never miss an issue of Stable Digest by subscribing for free!
Have suggestions for artists spotlights or community features? We wanna hear from YOU. Email us community@stability.ai
FOR THE COMMUNITY
StableChat/Code/LM AMA
Missed our Stable Stage discussion with the LLM Research Team? Catch the full video on our YouTube channel for all the inside scoop on StableChat/Code/LM.
Time’s running out! Join us at CogX x Stability.AI
You still have a chance to apply for an exclusive opportunity to join us at the CogX Festivalin London! You'll get access to amazing insights, captivating talks, and a chance to network with leading AI experts. Apply now.
BY THE COMMUNITY
Here at Stability, we're all about celebrating the vibrant AI community around us. So, we've created this little corner to share some of the fab things folks are creating out there! Just a heads-up though, just because we are showcasing these goodies does not mean they are all affiliated with us.
TOOLS & TUTS
Tutorial Spotlight!
Buzzing in the AI world recently is Gaussian Splatting, a game-changer for realistic 3D imagery, it uses a smart technique to take different pictures and blend them into a 3D model, adjusting itself to capture the fine details. We've curated a list of must-watch tutorials to get you started.
Check out The NeRF Guru's step-by-step guide on YouTube, where he walks you through install to creating your first scene.
And if you want a quick intro to the topic, don’t miss IndividualKex’s explainer video.
COOL TOOLS!
4MOOF
Ever wanted to know what it would be like to merge four images into one? Well, wonder no more. Artist @jeffjag has created a fun ‘little’ ComfyUI workflow that does just that. Grab it on Github and start playing today!
Reverie
@aviel08’s ReverieAI promises to be the next big thing in 3D environments. This advanced AI tool is designed to not just recognize and extract visual elements, but also categorize lighting and sound effects. Then, it seamlessly integrates them into your UE4 and UE5 environments. Eager to know more? The beta version is expected to launch later this year, so subscribe for the latest updates.
WITH THE COMMUNITY
STABLE SOCIETY DEEP DIVE - KEZIAI
“Be yourself; everyone else is already taken” - Oscar Wilde
KEZIAI, thank you so much for joining us today! Your journey is both inspiring and captivating—going from an award-winning director to pioneering AI artist, all while battling complex health issues. To kick things off, could you share the story of how you got to where you are today?
Thanks! It’s been a wild ride.
I was living in London and traveling the world, directing television commercials for clients such as HSBC, Cadbury, Mastercard, and Ponds. It was a busy life, and I was passionate about visual storytelling.
In 2011, I was shooting one of the biggest commercials of my career in South Africa, when I fell ill and never recovered. Later that year, I sustained a severe concussion, leading to vertigo and ongoing post concussion syndrome. Now back in Auckland, New Zealand, I’m 90-95% bed-bound, dealing with pain, fatigue, headaches, and cognitive dysfunction – all made worse by exertion. One of the most heartbreaking aspects was losing the ability to create, which had always been my raison d'être.
After years of what I call “life-grief“, AI transformed my life. I discovered AI art creation in July 2022, during a severe health crash, barely able to speak, move, sit up. Lying down with one eye open, I just had the energy to text. When I got the first images back my heart sang. I literally put my hand to my heart and felt hope, relief and excitement, revelation, liberation, and newfound opportunity. AI has enabled me, given me back my voice and rekindled my creative spirit. Prompting AI is like having a conversation with a friend, with beautiful art as the result.
Throughout my career I’ve always loved experimenting, and integrating technology into my work. Creating with AI feels like a culmination of everything I've done thus far. Forever curious, I have always loved learning, and pushing the boundaries of creativity, now from bed and learning in Discord back channels with boffins.
Combining my passion for concept, visual storytelling and technology, I’m excited about the possibilities of AI in photography, art, filmmaking, animation, 3D, and interactive art etc. I started filmmaking while at Elam School of Fine Arts in the 90s, studying Intermedia (time-based media) and also papers in Electronic Music, Art History, TV & Film, creating experimental videos, animations, short films, sound works, interactive DVDs, photography and installation art.
Since your shift from directing to creating AI-based art, how do you feel your background in storytelling has influenced your approach to creating AI art?
Creating with AI feels like directing in some ways, instead of directing a crew, you’re directing the AI. With both there is collaboration
Story can be told within a frame (looks between characters, action, the way your eye moves around the frame), in the moving frames of a film (time, dialogue, art direction), or in a curated series of stills (order, movement, colour: from one image to another). I love to play with all the elements of colour, character, emotion, movement, atmosphere, lens, style, wardrobe, composition, concept, novelty and story, whether creating with film, photography or AI.
After years of working with film and photography, I have a brain full of ideas. I love using my knowledge of genre, style, lenses, shots, art and film history, and various techniques to craft compelling prompts.
Filmmaking is expensive and hard work, and I’ve always wanted to be able to tap the imagination directly to create images and ideas as they emerge: AI taps the imagination.
The nature of my storytelling has also developed: for me filmmaking was mostly imagined worlds dreamt up that were thoroughly pre-imagined and meticulously planned. With my CHRONIC collection I am telling more of an internal story about the situation I am in, with the only tool possible. The process is pre-imagined but also going with the flow and investigating different approaches to promoting the idea.
The process of editing is similar to photography and film, you still have to ‘kill your darlings’, whatever the medium. With AI you create a lot of images on the way to the final, and curation can be brutal.
Speaking of your CHRONIC series, it evokes such an ethereal, dream-like quality, tinged with a sense of melancholy. Can you delve into how this series came about, were these pieces a therapeutic outlet for you in regards to your illness?
Sure. CHRONIC (A Revolution of Resilient Souls) is a collection of 150 images that were created over seven months from bed, about and while experiencing differing states of illness: head feeling inflamed and on fire, eyes feeling pulled back inside my head and from the pull grows flowers, lying in bed in a fog, discombobulated in a vice etc. Curated from thousands with the help of the wonderful Jacqui Kenny of Argleton Lane.
I want to create work that raises awareness for the millions suffering from invisible illnesses, but that doesn't make people want to look away, ‘CHRONIC’ is a very personal collection but also universal to the human condition. It has been therapeutic to create.
From Top Left to Bottom Right - CHRONIC #24, #98, #139, & #126 by KEZIAI
Melancholic threads entwine with hope's embrace, grotesque dances with beauty
Metaphorical, surreal images transcend time and bend reality
Lost in the fog, tangled in hair and sheets, head ablaze, trapped in a vice
Coding language with AI, chronic realities find solace in creative dreams
CHRONIC is also available to view in the Gallery in the Stars on The Line.
Other artworks, such as ‘I'm Fine’ and 'You Look Fine’, also address the theme of invisible illness, highlighting the disconnection between outward appearances and the harsh reality experienced by individuals who often hear the well-intentioned but misleading comment “You look fine”, even when they may be in extreme discomfort, exhaustion, or distress.
‘You Look Fine’ - KEZIAI
Your work and story have resonated with a community of artists and enthusiasts who also face various health challenges. Can you speak to the role and importance of AI art tools in making art more accessible for differently abled creators?
AI art tools have played a significant role in making art more accessible for differently abled creators, including those facing various health challenges. These tools have fostered a sense of community among artists and enthusiasts.
The availability of AI art tools on mobile devices has been a game-changer, enabling artists to create from their own space, even lying down.
The more I speak about the inclusive aspect of AI, the more creators I connect with who also have been empowered by these tools.
While there will always be skepticism to new technology, AI art is just another tool, and like any tool, it requires effort and dedication to master. In my specific context as an artist, AI art has been brilliant and has provided a new dimension of creative possibilities.
One intriguing aspect of your work is how you invite collectors to participate in shaping the narrative, even naming characters. What inspired you to bring this interactive element into your art?
‘Bubblegum Queen’ is a fairytale within my first AI collection, ‘Fairytale Dreams’, August ‘22. The inspiration for this fairytale was the trait gangs the Aipes NFT community had formed around their AI PFP collection by Kin. There was the Flipped Brim Gang, Gold Gang, Bubblegum Gang and I created AI characters within these. You can buy a character and name and describe it and I’ll add that info to the metadata, or you can buy a ‘blank flag’ of the Gang you want to be part of. You then think up a character and brief me with the name and description and I’ll imagine it up with AI, giving you several options to choose from. The story unfolds to include location stills for Ethereum Forest, Bubblegum Castle and storyboard frames as the scene is set.
BubbleYum & Flipped Gang - KEZIAI
Additionally, using Chat GPT collectors can write a unique fairytale inputting their character’s traits (name, family, gang, role etc) into the blanks left in a fairytale-creating prompt I provide. I love the idea of using the nature of the technology to inform the artwork, and NFTs seem perfect for this type of interactive storytelling.
Another fairytale, Flying Bed, features flying beds in a fairytale world. When I was a kid I invented a flying bed with all the bells and whistles, to fly away when sick off of school. Now with AI I can create it!
I’m now also interested in dynamic NFTS and the potential for interactive involvement.
Many artists struggle with the notion of perfection. In your artistic process, you work closely with AI, iterating until you are satisfied with the image. How do you know when a piece is complete, especially when working with a tool that can endlessly generate new possibilities?
Good question. When you can make anything, what do you make and when do you stop iterating? I create numerous images when aiming for an idea, sometimes the more specific image you have in mind the more prompt-crafting and setting tweaking you’ll do. Sometimes I get what I’m after quickly but more often than not I try many different iterations and different ways of looking into the subject. I like to do most of the creation with prompts and settings and use Lightroom for tone and color; but this means making many versions and a lengthy selection process.
I use little to no Photoshop on most of my pieces, however some do use Photoshop, such as “You Look Fine” (generative fill and comping) and “Touch Grass” (comping three images together).
When the tuning fork stops is when it’s complete: when the image says the message you want to convey and the visual feels complete. When my eye no longer gets ‘caught’ on something in the image: when it’s working, the eye moves freely around the composition.
Also, I find deadlines frustrating but helpful. LOL.
The rise of AI art challenges traditional definitions of what constitutes "art." In your opinion, what are the core elements that make something worthy of being called art, irrespective of the medium?
"It's art if you say it's art" - commonly attributed to the artist Marcel Duchamp, emphasising the role of artistic intention in defining what is considered art.
Humans have always made art with what they have around them, self-expression being a component, pattern making is in our genes, whatever the medium. I’ve tried creating with most mediums at some point and they are all a dialogue between the idea and the medium, between the hand, the head, the eye and the medium. Likewise with AI.
To me, AI is simply another medium, and the intent, act of creation and curation make it art. Prompt-crafting, techniques and style add uniqueness and context.
Whenever there is new tech, its validity as ‘art’ gets questioned. Happened with photography, with Photoshop. It also happens with new styles: the Surrealists faced opposition and criticism from the traditional art world due to their unconventional and provocative artistic style and ideas.
You’ve mentioned that the Web3 world has introduced you to a large, like-minded community, many of whom also face health challenges. Can you speak about how this virtual community has impacted you emotionally and artistically?
Web 3 has been incredible. Finding like-minded curious humans interested in ideas, tech, art, film, music, health all together in this space is very cool. When I’m able to, energy wise, I love connecting on spaces and getting to know artists and collectors from around the world. It’s been a very supportive space as I gradually came out about being ill, which had been private until last year.
Artistically I love creative energy and opportunities for exhibition curations and open calls, encouraging experimenting. Being in group chats, DMs and X (formerly Twitter) Spaces has helped sharpen my intent.
Your award-winning short film 'Nurse Me' is a testament to your storytelling prowess, so it's exciting to see you venture into AI animation. Could you tell us a little bit about your recent projects like ‘Into the Ether/Liminal Reverie’? Additionally, how do you see the relationship between AI and cinematic storytelling evolving in your future work?
Thank you! ‘Nurse Me’ is a poetic dance film exploring bodies, architecture, sensuality and existence.
I’m excited by the possibilities of AI moving-image: I’ve created with various tools and techniques. Into the Ether/Liminal Reverie draws inspiration from a personal near-death experience 20 years ago, where it felt like cells lifted off my body and dissolved into the ether; this poetic video artwork is a visual exploration of the boundary between life and the ethereal. A lucid dream... I’ve wanted to visualise this experience, but until now haven’t had the right medium to imagine it.
Into the Ether/Liminal Reverie (sound on) - KEZIAI
I loved creating ‘(e)motion’ an AI animation created using Stable Diffusion, Automatic 1111 with Deforum, Parseq and ControlNet, testing every parameter to get a smooth 3D camera movement through emotions, using a Midjourney init image.
I’m excited about evolving AI tech and cinematic storytelling, both as an aid to traditional filmmaking but also, in my context, enabling me to be a filmmaker again, which I thought wasn’t possible. What you can create now I thought was several years off, so where will we be in a year!
(e)motion (sound on) - KEZIAI
Thank you for taking the time to speak with us! Would you like to share any additional thoughts with us and our community of readers, or give shoutouts to anyone who has been an inspiration to you along the way?
So many! Especially Jacqui Kenny for her support, curation skills, friendship and talent. I was pleased to be her first collector and now she is flying. Her collection Argleton Lane is an imaginary world based on Google Earth images Jacqui took while touring the world from home, with Agoraphobia.
Follow me on twitter @keziabarnett for updates, and follow @keziai on Foundation for minted work. I’m currently prompting every day of 2023 with a formulae prompt, which serendipitously documents the evolving versions and models as they change.
“Be yourself; everyone else is already taken” - Oscar Wilde
Through the Ages Winners
We are excited to announce the winners of our Through the Ages contest! This was the first of a variety of contests and events happening in our discord server to celebrate Stable Diffusion’s one-year anniversary!
1st Place - Konsequencetoyz
2nd place - TophatPandaMagician
3rd place - humblemikey
And just a reminder, the party’s not over! Be sure to head on over to the announcements & events channels on our discord to check out the other giveaways, Stable-Stages, and creative contests!
WINNERS OF THE WEEK
The COW prize goes to Minkuhh, for their adorable rendition of Beyond the Toybox that is sure to spark a little warmth into anyone's heart!
Minkuhh - Beyond the Toybox
Head on over to the Community Events on the Discord and earn your chance to be highlighted!
