Dear M○C△: What Art Excites You Today?
In so many words...
Please enjoy today’s DEAR MOCA column. Have questions you’d like answered? Have a thought you want us to respond to? Email us at hello@museumofcryptoart.com or submit your questions to our dedicated Google Form.
Welcome one and all to the long-awaited return of:
And this week, we have a good one:
Dear MOCA,
Really open-ended art question here: What kind of art is exciting to you today?
Sincerely,
Aesthetically Minded
Dear Aesthetically Minded,
Initially saw your question and thought “Man, this is going to be eeeeasy.” But then when I sat down to actually put words to page, I realized, “Man, this is going to be haaaard.” I’m much better at chronicling history than identifying current trends, so like anyone who gambles in memecoins knows, things that seem obvious in hindsight never seem that way in the moment.
A few general trends do seem to be capturing public attention, however, and where attention goes, artists will en masse follow, so perhaps innovation and excitement is soon to come. The aesthetics across-the-board leave a lot to be desired, in my opinion, but this XCOPY-bolstered resurrection of the Open Edition meta is at least worth mention. The Open Editions themselves on a Base-centric market mechanism and not an aesthetic or artistic trend (and one which I assume will be fairly short-lived), but it does does seem to be rejuvenating interest in 3D art, much of it paying homage to NeonGlitch86 and XCOPY’s Mutatio (Mew-tate-ee-o? Mew-taysh-o? Mew-taysh-a?), the 3D fly with its little Xed-out eyes and million-minted editions.
On top of that, you have Sovrn.art’s Cent project featuring 10,000 archived pennies (literal pennies) minted as Bitcoin Ordinals, which as a commentary on value and inflationary Fiat currency and the nature of having all pennies made of copper (pre-1984 mints) being worth $.01 in economic value but worth $.025 in raw material, like, that’s super interesting, and I hope this work does well, and I hope it sparks a ton of conversation, but conceptual art is fairly difficult to emulate, and I can’t point to really any other recent Ordinals projects as revolutionary (though I’d love to be educated by anyone who disagrees). That ecosystem itself seems much more speculative than artistically interesting, so good on Sovrn.art, but the excitement might start and stop with them.
Ultimately, my attention for months has been eclipsed by a single submovement, it’s on my mind basically all the time, it’s the subject of many MOCA podcasts recently, it energizes me as much as it defeats me, mesmerizes me, frightens me outright, and that’s AI art.
We pay a lot of attention to AI, and I’m still not sure we’re paying as much as we should be. And I’m not talking about AI art in terms of raw aesthetics —which, for the record, have become routinely astounding— but all its earthshaking new application. We have been saying for a long time that AI is going to fundamentally redefine every art form on Earth, but god damn it, we’re here. The final frontiers are currently being crossed, and I wanted to spend some time today talking about three specific projects that I find especially monumental, and why, and what their impacts might be on, well, every single thing we do, say, think, and believe.
NousResearch’s WorldSim
Okay, so I was just exposed to this thing like 2 hours ago, but it’s already blowing my mind. For those of you who aren’t familiar with WorldSim’s creators, NousResearch is a multi-nodal open-source AI research organization. They do a great deal of radically-cool shit. Their work, in particular, focuses on Large Language Models (LLMs), which are the kinds of AIs made famous by OpenAI, but Nous also do a great deal of ecosystem incentivization, rewarding open-source developers for creating better, quicker, more accurate, more sophisticated models. I know some of the folks working on NousResearch, like the artist and AI researcher Karan4d (who I’ve written about at length elsewhere), and they’re good folks, brilliant folks, and everything they do is brilliant, but this WorldSim thing…this is something else.
The scope of WorldSim is almost hard to describe, but from what I understand, you provide the model your description of a literal universe. Like anything you want. For this example, I’m going to write (off the top of my head), “A universe where the fundamental building blocks of life are pulled towards artistic formations, where gravitational force is replaced by creative chemical composition, and where complex organisms emerge from seeds of creativity instead of organic material.” Let’s see what it says.
The model —I have zero conception of how you train something like this— spits back the very atomic organization principles of this completely ridiculous, slapdash universe I’ve put together.
In this universe:
“Matter now interacts via creative principles instead of spacetime curvature.
“Particles arrange into aesthetically resonant structures and patterns.”
The model has populated this universe with “Planetoids composed of pigments and clays formed, surfaces etched with frescos.” It even imparted narrative onto the thing, saying that “Critics and Curators merged via empathic abstraction, experiencing artworks as from every perspective at once, an omniscient Creative Overmind. It beheld the elemental beauty underpinning All, and in that moment, contemplated escaping the confines of physicality altogether…”
WorldSim calls the universe “Aesthetica.” I couldn’t have done a better job myself.
Let me tell you, as a fiction writer who everyday faces the tremendous challenge of inventing a new world from scratch, this is something far beyond revolutionary. This model is capable —already, today, on Day 1 of release— of creating universes at or beyond the sheer scope of human comprehension. A nuance I’ve always noticed with sci-fi, take Star Trek for example, is that humans can imagine the advancement of certain technologies —computers, phones, weaponry, space travel— but it’s exceedingly difficult to extrapolate how that technology develops exponentially and upon itself. Thus most versions of the future we see are really just slightly-tweaked, slightly-advanced versions of Earthbound human reality. Ursula K. Leguin, one of my favorite writers, has come as close as anyone to reimagining societies comprehensively from the ground-up, but I guess all that striving and struggling is now bunk, because this model can basically do the most impossible part of the work on-demand, in literal seconds.
I asked the model “What kinds of life-forms are most common here,” in a universe invented specifically for the purpose of this weekly column, and it revealed something so achingly beautiful I feel both inspired and creatively devastated.
I’ll have more superlatives to talk about with WorldSim in the future, but for now, my brain has been scrambled, served on a paper plate, and appears to have been doused in hot sauce, so as I go take a shower or something, I’ll just let the magnitude of the thing speak for itself.
Kojii.ai
Full disclosure, I’ve done a bit of writing for Kojii.ai, and I’m quite close with some of the team-members, so before you think “Hey, pretty suspicious of this dude to be out here aggrandizing a product he has even slight financial stake in,” (which I don’t, to be clear, I just wrote some profiles at their request), I wanted to be up-front.
But I want to stress, I don’t just go around writing for whomever offers me some cash. I tend to keep my talents reserved for folks and projects I believe in. Pindar van Arman’s Monograph, RightClickSave; when I leave MOCA’s orbit, it’s for stuff I believe in. And I believe in kojii.ai as not only a successful product, but as a sea-change for how we even conceptualize AI art.
The gist of kojii.ai is that they enlist artists to train AI models on a some specific aesthetic or conceptual vision (using Eden.ai’s back-end tech), which the general public can then use in-browser to generate artworks within the initial conceit. They describe this as a “playground.” It feels exactly like a playground.
The models themselves exist in the middle-ground between generalized, and easy-to-use prompt generation models like Dall-E or Midjourney and the intricately-trained StableDiffusion models that, yes, many of the top AI artists use, but which require immense amounts of technical skill and computing power to download, let alone use efficiently. Every one of these projects has its own thesis, underlying concept, tremendous dataset, and bespoke UI/UX design with customizable parameters. All of which, again, are designed by the artist responsible from the bottom-up. Kojii.ai offers access to these artist-designed models on demand, and then you can mint the works you generate, if you so choose.
To me, this just screams “FUTURE!” The public’s wariness of AI artistry stems from the perception that AI takes all the skill out of artistry, which is itself completely untrue, but is based on reasonably firm logic. I don’t have to be a trained painter to turn the sentence “Cherry-blossom trees by a secluded lake in Northwest Seattle in the style of Monet” into a fairly gorgeous Impressionist painting. But Kojii.ai redirects attention away from the individual outputs and onto the models themselves as the artwork. The grand glut of the generations is the artwork, not any output singularly. Every model is essentially a high-concept performance piece, requiring the public’s participation, who then get to ingratiate themselves and their own desires together within the conceptual parameters of the artist. From this emerges a synergistic artistry that could only ever exist from specific choices, in that specific moment, by those two minds.
Imagine, in a few years, we might be socioculturally accustomed to a version of AI artistry which is not necessarily connected to outputs at all, but to the precise, conceptual, individualized models themselves. The artistry is transplanted from the creative product onto the creative process itself, and that creative process can be packaged, ported around the world, used by whomever, and then ultimately marketed in a way which benefits both the user (who mints and thus may sell the work), and the artist who designed it.
I can’t see this stopping. I can’t see it slowing down. I see this becoming the new normal. I’ll give you six months-or-so before I come around handing out I-told-you-so’s
OpenAI’s Sora
This one isn’t this too complex on it’s face. The great artist Makeitrad was telling me how, when he first started exploring AI artistry, the problem of animating outputs —highly desirable— was insurmountable, though some creative workarounds eventually appeared. He described to me a process where the AI would generate static images as connected keyframes, so you could approximate motion as one would with a flipbook. The technology has developed further since then, but OpenAI’s coming prompt-based revolution in video and animated AI artistry is about to supersede every obstacle in the film, advertising, television, streaming, and videography industries. Probably others too, which I can’t currently conceive.
If you haven’t seen the kinds of videos OpenAI’s Sora can generate, do yourself a favor and spend ten minutes educating yourself.
When AI can be compellingly used to create narrative artistry, which generally requires the sequencing of frames, I promise that the public penetration will be staggering. Within a month at most, we will see short films, fantastically compelling short films, generated and audio-laden entirely with AI. And these videos will make people laugh, they will make others sob, and another bastion of “Only humans can ever do this,” will fall. It will fall quickly.
Video is the predominant communication medium of the most predominantly used social media networks for people under 30. Youtube, Tiktok, Snapchat, and Instagram all have video at the nucleus of their platforms. Replace merely 10% of those videos with professionally-conjured, on-demand AI generations from Sora, and the entire fabric of public communication shifts unalterably.
The artistry that is going to come out of Sora will freakin’ rule, by the way. Let the brilliant AI artists in crypto art get their hands on ready-made narrative-making tools, stories they can tell by using their brilliance alone as currency, and we’re in for some real treats.
So anyways, that’s a long way of saying, yes, I guess I’m excited about quite a few things in art after all. Who would’ve guessed?
- Your Friendly Neighborhood Digital Art Museum’s Friendly Neighborhood Digital Art Writer,
Max from M○C△