13 Aug 2024/62 min/Chikai
MLow x Chikai Interview
Collector turned artist Chikai on Circle of Friends, defining cinematic AI, and his openly ridiculous goal of making the first AI film to win an Oscar.
Chikai Ohazama and MLow open by admitting they have only had flybys at events, and MLow explains why the show exists at all: Twitter Spaces are uncontrolled environments full of distractions, and you only reach the heart of a conversation one on one. The two mirror each other, both collectors before artists, and Chikai's arc was entirely unplanned. Circle of Friends began as homage to artists he could never afford, like Sam Spratt, referencing their color palettes rather than their work. He minted the first season on January 1st purely because OpenSea had promised to grandfather royalty terms for contracts made before January 2nd (a promise later reneged on), planned to check back in a few days, and watched it mint out in minutes, assuming at first he had been hacked. Season two jumped from 500 pieces and 8 artists to 10,000 pieces and 25 artists because nearly every artist he asked said yes, teaching him that selling 10,000 of anything is non-trivial.
The middle of the conversation belongs to two of the archive's best formulations. Marketing, in Chikai's telling, is not selling but sharing a truth you genuinely believe and finding the most authentic medium to convey it; for an artist, the truth is you, and if he had needed to be anyone but himself he would have failed miserably. And collecting has changed: buyers now expect connection, community and access as part of the purchase, so its absence feels like a defective piece. MLow adds his own experience co-owning the second to last Squiggle with about 3,000 others through Snowfro's final mint event, and they discuss museums like LACMA taking donated works as institutions gain skin in the game.
The final stretch defines cinematic AI, the genre Chikai works in and named for himself: born the moment Runway shipped public image to video, distinct from earlier AI animation, and consciously built on cinematic craft from Miyazaki to Wes Anderson, as against the Bill Viola video art lineage. He maps three worlds that barely touch (Silicon Valley tech, NFTs and Hollywood), argues AI is already inside Hollywood through concept art, the Everything Everywhere All At Once effects team, and AI marketing posters for Civil War, and lays out his practice: share 99 percent of what you make moments after making it, stay prolific because ten average scripts teach more than one perfect one, and keep one unfathomable goal, being the first AI film ever to win an Oscar.
What this episode teaches
- You have to jump in
Chikai says web3 cannot be learned by reading: you just have to try it and figure it out, hopefully without getting hacked, drained or scammed along the way. The space also tolerates multiple hats, so a collector can become an artist without being told to stay in a lane.
- Marketing is sharing a truth
His reframe: marketing is not selling, it is sharing a truth you genuinely believe and getting someone else to believe it. For an artist the truth is yourself, so the whole craft reduces to presenting yourself authentically and picking the medium, posts, video, podcast or just the art, where that truth lands.
- Homage lifts peers instead of competing
Circle of Friends made inaccessible artists collectable by referencing their color palettes rather than reusing their art, and lesser known participants received waves of inquiries afterward. A model for supporting artists you love while building your own practice.
- Selling 10,000 of anything is non-trivial
Scaling season two to 10,000 pieces because 25 of 26 invited artists said yes taught him that distributing ten thousand of anything, stickers, shirts or NFTs, is a serious operational undertaking. He adjusted prices and mechanics and still marvels that it sold out.
- One peer's comment can start a career
ACK saying he liked one of Chikai's early AI pieces is the entire reason he found the confidence to mint his first work. Encouragement is cheap for the giver and decisive for the receiver.
- Vision first, tools in service
His advice to artists adopting AI: never be driven by the tool. Have a clear vision of what you want to say, then pull in AI, CGI or anything else to realize it. Most current AI work is eye candy showing off the tool with no vision behind it.
- Prolific beats perfect
Getting stuck polishing your first masterpiece stops growth; the person who writes ten average scripts develops faster than the one who rewrites a single script. The most successful artists in any form are incredibly prolific, and most of their output never reaches the album or the auction house.
- Keep one unfathomable goal
His dream of being the first AI film to win an Oscar sounds ridiculous, and he says that is the point: audacious goals that seem stupid give you something industry changing to strive for, and even reaching a fraction of one is success.
The conversation, topic by topic
Multiple hats and learning by doing
Chikai notes the space accepts people wearing collector, artist and builder hats at once, and is grateful his artist side was accepted by peers. His one firm lesson about web3: you cannot learn it from reading, you have to jump in, and MLow calls it burn to learn. MLow also explains the show's format: one on one recorded conversations reach depths Twitter Spaces cannot.
Circle of Friends and the accidental sellout
The collection began as a way to own a little bit of artists he could never afford, like Sam Spratt, via palette homage rather than their actual art, extending his Monolith Gallery instinct for visibility. He minted on January 1st only because OpenSea said contracts before January 2nd would be grandfathered for royalties; it minted out in minutes before he could even reserve pieces for the honored artists, and he briefly assumed he had been hacked. Then it kept climbing on secondary.
Scaling from 500 to 10,000
Season one was 500 pieces with 8 artists. For season two he invited about 26 artists expecting half to respond; 25 accepted enthusiastically, so he raised supply to 10,000 and discovered selling ten thousand of anything is real work, selling out only with momentum from DeeZe and others.
The collector network and ACK's push
Two gifts from having been a collector first: peers like Gee and NorCal accepted his artist turn, and he understands the collector's side of outreach, including that silence usually means busy, not rejection. He researched AI art to understand what he was buying, started creating during the research, and minted his first piece only after ACK complimented one.
Truth, authenticity and the accidental personality
His marketing philosophy is sharing a truth you believe; inauthentic marketing is just a tool and fails. He never planned to become a Twitter personality, finds words like influencer counter to who he is, and accepts the role as a blessing and a curse because of the good it enables. If he had needed to be anybody else, he would have failed miserably.
The Squiggle mint and buying connection
MLow recounts sending ETH to Snowfro's final Squiggle event, getting it returned, and later learning he co-owns the second to last Squiggle with about 3,000 people through the Snowfro and DEAFBEEF collaboration. Chikai agrees connection is now part of the purchase: when he buys art without feeling part of a community, something feels missing, an expectation that barely existed for the average art buyer historically.
Museums and the new value system
They discuss Snowfro donating works to institutions like LACMA and SFMOMA (where Alejandro Cartagena is involved), giving museums skin in the game. Chikai wonders what this era's MoMA equivalent will be, whether high signal individual collectors or a native institution, and raises networked art, ownership spread across hundreds of collectors, as a genuinely new idea, noting crypto art's whole history is barely three years old in its current form.
Defining cinematic AI
He dates the genre to a single event: Runway launching public image to video, the first time he could see what was in his head become a piece of art. Cinematic AI is his sub genre within animated AI art, defined both by the image to video technology and by borrowing cinematic craft; his influences run from Miyazaki to Wes Anderson, and he distinguishes the lineage from video art in the Bill Viola tradition. The name marks which history you are building on.
Image to video versus text to video
He finds image to video most expressive because text to image control (Midjourney) is far more developed, while text to video, despite dramatic recent progress from Sora, Luma and Runway, offers less variety and control. Generation speed matters too: a creative flywheel needs low latency. MLow shares his own workflow, using Luma after early Runway work, and jokes about how far AI video has come since Will Smith eating spaghetti.
Three worlds that barely touch
From his Silicon Valley career (MLow notes he founded Google Earth), Chikai observes he was nearly the only one in his tech network to go deep on NFTs; companies like Luma barely know NFTs exist, focusing on Hollywood and creator platforms. Even Andreessen Horowitz splits internally between Chris Dixon's crypto cohort and the AI investments. He maps Silicon Valley, crypto and Hollywood as three separate worlds whose eventual convergence will bring destruction and creation at once, and he hopes to connect all three.
A once in a lifetime cohort
He compares the COVID era gathering of artists and technologists online to the Renaissance, Warhol's Factory, and Gertrude Stein's salon hosting Picasso and Dali: a place that was the internet, everywhere yet nowhere. Being able to text artists he believes will be historically renowned still amazes him, and he calls this cohort of artists and collectors a once in a lifetime one.
AI is already inside Hollywood
He argues Hollywood adoption is further along than the polarized debate suggests: the Everything Everywhere All At Once effects team used Runway tools for masking, producer Brian Grazer cited AI and CGI for an upcoming sequel, Civil War used AI generated marketing posters, a Marvel show ran an AI generated intro, and pre-production concept work is quietly being done with AI at a fraction of former cost. He believes the industry will have to jump on the bandwagon.
Democratized creativity and the story bar
The biggest change he expects from AI video: a greater share of the population can manifest what is in their heads without a director's budget, and skills like film editing become practicable by anyone. Because pretty pictures are now trivially easy, quality must move up a level; he hopes mediocre shows become unacceptable, citing Nintendo beating Xbox and PlayStation graphics power with gameplay. The secret sauce stays the same: have a story to tell. He also names AI's most personal use, getting unstuck, describing how generated dialogue rescued a Wes Anderson style short that had stalled for weeks.
What's next: festivals and the Oscar dream
Cinematic AI 2024 launches within weeks of the recording. In parallel he has spent a year making short films and submitting to every major festival, with one acceptance so far, and openly holds the dream goal of the first AI film to win an Oscar. His 500 piece BrainDrops collection, which he believes is the largest cinematic AI collection ever assembled at that scale, showed him how much tedious work video coherence at 500 pieces demands. He closes by crediting MLow's genuine, authentic way of connecting with people as the reason guests trust the show.
In their words
“You just got to jump in and figure it out, and hopefully don't get hacked or drained or scammed along the way.”
Chikai“Selling 10,000 of anything, like stickers, t-shirts, whatever, pick anything, is non-trivial.”
Chikai“You're trying to share a truth that you truly believe in with somebody else, and to get them to believe that truth.”
Chikai“I somehow was able to just be me, and people liked it. If I had to be anybody else, I would have failed miserably.”
Chikai“The most successful artists in any form tend to be very, very prolific.”
Chikai“This is a dream goal which may never ever happen: to be the first AI film to ever win an Oscar.”
ChikaiCite this conversation
Chikai in conversation with MLow, "MLow x Chikai Interview," The MLow Show, 13 Aug 2024. https://show.mlow.xyz/e/2024-08-13-mlow-x-chikai-interview.html. Quoted from a machine transcript.
Quotations come from machine transcripts. Check wording against the audio before republishing.