Joshua Larson

@godatplay ∙ Published June 16, 2023

What’s your background and how did you get into AI art?

My first real job was working at a mom and pop clothing store and learning about screenprinting and graphic design for apparel. I later went to art school, so I’m classically trained as an artist and designer, but I eventually got into digital media, starting with web design, animation, and 3D. I eventually gravitated toward generative art - the kind where you write code to create art - and game development, both of which I’ve been doing for about 20 years.

In 2022, I was in a deep depression and lost my creative spark, both for the first time. It was very unfamiliar. I didn’t feel like developing games, or creating generative art, but I could at least type some text prompts into this new tool Midjourney. Just two weeks into using that, I had a transformative experience that helped to reignite my creative spark. That moment made me realize I needed to dive head first into generative AI because it would become a major tool in all creative disciplines. I became an expert in it, or as much as one could at the time, and while creating character concepts for a game I was directing, I realized what I loved about working on the characters was that I could be a fashion designer. This was my chance to do more with fashion like I always wanted since my first job.

What drives you to create?

In short, it’s just an innate need I have. If I were to guess and be more specific, I think I love the process of having an idea and thinking about what value it would have, and then translate that weird mysterious blob of idea into a lump of clay in front of me that I can shape. I also love to get feedback on my creations, especially critique.

What does your workflow look like?

Identity is usually where I start when I go to create. In videogames, when you create an entire world from scratch, you need to decide its history, its sense of right and wrong, and how players can express themselves and make choices. Fashion is ultimately about identity, too, so I appreciate how I can start in the same place.

My environment has a big impact on my workflow. Two communities that influence my work are my church and my studio at Mainframe Studios. They’re both big melting pots of people from different races, socioeconomic backgrounds, artistic media, and birthplaces. My experience living within these communities gives me a lens that affects everything I do.

Through that lens, while reflecting on identity, lately I arrive at futurism. And not like the movement in decades past that cheered on destruction and derided women. My futurism is one of hope and a multi-cultural celebration.

From there, I’ll experiment, iterate on ideas and implementations, refine concepts, and then at that point I usually have to ship, even though I don’t feel like I’m ready! With creative brainstorming and even general design, Midjourney is a major AI tool I use. Sometimes I’ve used Stable Diffusion to create my own custom finetunes, but I often want more coherence, so I’ll be using it more heavily if I can use a newer model as easy to finetune with as 1.5.

What is your favourite prompt when creating art?

As a creator, a part about the process that feels particularly magical is when you refine a single prompt that generates a cohesive collection that matches your goals and you think will connect with the viewer. Because of that, for AI image generation that is not heavily modified, I personally think the art is in the entire collection, which is to say I think the art is in the prompt itself. If an artist is using an AI-assisted process to execute one image they’ve imagined ahead of time, it’d make sense that their art would be a single image. For everyone else, I wish AI art would move in this direction.

A prompt phrase I like to use is hip hop haute couture. Depending on how I add to it, it can go in many directions. Here’s some example work:

I often use a few images as a prompt as well, so I may take several results from prompts using that phrase, and use them only as image prompts. Over time, I find images that are special, they pass on interesting traits to their genetic offspring as it were. Here’s some example work using a couple of those special parents:

How do you imagine AI (art) will be impacting society in the near future?

I think AI will change every creative industry. I can best illustrate that with my own story.

Two weeks into using AI, I had a transformative experience. In June 2022, I was trying to recreate the architecture for a VR game I was directing. One particular prompt I thought of resulted in this new form of architecture I’d never seen before. What blew me away was that I had already tried this general direction and written it off as not good enough, but what I saw was that same direction but with a wild twist.

The classical creative process is to create a lot of rough sketches, throw out the terrible ideas, start another iteration with more refinement and curation, and then repeat. The whole time you have to extrapolate what a finished creation will look like. I realized there was no way to conceive of this wild twist, and as a result, there is an entire possibility space of amazing art and design that will never be explored through the traditional means, because creative people have to curate before the work is finished.

The only way to find all these potential creations is to simply not care whether or not something will work or not and curate at the end instead. And the only way that new process makes sense is to make the process of creation incredibly fast.

Who is your favourite artist?

• Traditional art: Modernist art movements inspire me in general.
• Generative art: Jared Tarbell and Joshua Davis got me into generative art in 2003. Matt DesLauriers and Arsiliath are newer examples. Thomas Wilfred is a more multimedia example.
• Videogames: Jeryce Dianingana, Ryan Green, Amy Green, Stephen Lavelle, Gregory Avery-Weir, Masayoshi Kikuchi and Ryuta Ueda and the rest of Smilebit, Jordan Magnuson, and David Kanaga.
• AI art: Đy Mokomi, whatisiana, Simon Lavi, and Sasha Stiles.

Editorial note: Check out the interview with Đy Mokomi here.

• Photography: Summer Wagner.
• Animation: Mirai Mizue.
• Fashion: Walter Van Beirendonck, Sekure D.
• Film: Andrei Tarkovsky, Terrence Malick.

Do you have a specific project you’re currently working on? What is it?

At this hour, I’m working on an alternate reality game entitled Forgotten Skin, which is a heist game that takes place across web3 and has a fashion focus to it. To start playing the game, you collect an avatar NFT.

“Kwame of the Eshun”, my piece for the AI Surrealism show, is one such avatar. Players are racing each other at all hours of the day to get access to one of a small collection of 16 ordinals - NFTs on Bitcoin - that were within the first 100k inscriptions out of the current 10+ million.

The game is a sort of prequel to a larger videogame that has been in development for a few years, and avatars from Forgotten Skin will be reused in this future game. The hour is at hand.

Would you tell us about the AI Surrealism exhibition you are a part of?

I’m really honored to be part of AI Surrealism. I’m in great company with the 99 other AI artists, thanks to the expert curation of Anna Dart and Roger Haus. They and the rest of the Exquisite Workers community have done a great job with the event.

I’m also thrilled I get to be part of a Superchief Gallery NFT show with the gallery aspect of the exhibition at Canvas 3.0 at the Oculus building within the new World Trade Center in NYC. I was able to visit there multiple times over its construction and never would have thought I’d have work presented there.

My work in AI Surrealism is “Kwame of the Eshun”, a fashion concept that can be used as a digital avatar for Forgotten Skin and a future game. This concept is part of an aesthetic I’ve been developing for about 9 months, which I liken to an Olympics of fashion. I want to celebrate the diversity of humanity, both how we’re fundamentally alike and how our cultures differ in a beautiful but harmonious way.

What does it mean to be an AI Surrealist for you in the times we live in?

I personally think AI image generation and surrealism go hand-in-hand perfectly, especially with the current state of the art. Breton referred to surrealism as a process where artists suppress their conscious mind and explore their subconscious to bypass any potential repressive states. To me, AI image generation is a kind of psychic automatism that encompasses all of humanity. The consciousness of humanity is encoded and then repressed, and then the artist works to bypass that and draw a specific element of that consciousness out based on their own unique identity.

Anything else to add?

Beneath the shroud of time,
a canvas lost and forlorn,
Forgotten skin,
a vessel where stories are born.

Ghostly echoes of past lives,
fading with a sigh,
Aching to be remembered
before they say goodbye.