Art and automation's merger long predates cryptoart's use of procedural generation.
You'll never hear NFT sellers talk seriously about that history, though, cause it reveals not just NFT art's contradictions, but also its cynical laziness.
I love Rube Goldberg contraptions. As a kid I was obsessed with Wallace and Gromit, with the strange machines in Pee Wee's Big Adventure, and with the nonsense cartoon inventions of Betty Boop's Professor Grampy. A Rube Goldberg machine is beautiful because of its pointlessness. It's a supremely inefficient and mesmerizingly intricate way of accomplishing something. As the kind of autistic woman that can stare slack jawed at reflections on water or gears turning for long minutes at a time, Rube Goldberg contraptions always tickled some deep part of my brain.
And so, later in life I got involved in procedural generation.
Procgen is a Rube Goldberg contraption par excellence. There's a saying among developers: procgen is a great way to do twice the work in double the time. Sure, there's lots of noise about it enabling expansive, always unique games that can be produced for a fraction of the cost of authored content. But I think procedural generation shines best with an aesthetic refocus: from a goal or aesthetic endpoint to the beauty of the process of creation, and the relationship between the two.
That's how it functions in fine art. Rube Goldberg machines found their way fully into art a century ago. Through the 19th and early 20th centuries, lots of artists moved away from final product and toward process. They emphasized not just an aesthetic or didactic result but the way those results were derived: rapid open air painting, random selection of "readymades", exploration of paint splatters and splashes on colossal canvases, performances that incorporated audience response or explored the limits of an artist's endurance. The Surrealists placed special emphasis on this, employing all kinds of tricks to get out of their conscious mind and tap into the unconscious. Take Frottage, for example, which among other things (ladies...) means taking charcoal rubbings of textured surfaces, which artists like Max Ernst then interpreted into strange phantasmagoric beings and landscapes. Or Dadaist Jean Arp's collage canvases of paper scraps dropped at random and then pasted down where they fell.
This is in its own way a Rube Goldberg way of achieving art, or to put it another way, it's a procedural generation, creation focused on method rather than a preset and defined end goal.
Much contemporary art, in fact much art since the 1950s, consists of methods and instructions--algorithms, really--that can be set in motion by ANYONE with the ability to perform them, much like a musical score. One of my favorite examples of this is when hip hop artists clipping. capped a horrorcore album off with a performance of Annea Lockwood's "Piano Burning." The score of the piece is:
Set upright piano (not a grand) in an open space with the lid closed.
Spill a little lighter fluid on a twist of paper and place inside, near the pedals.
Light it.
Balloons may be stapled to the piano.
Play whatever pleases you for as long as you can.
The score gives quite a lot of leeway to adapt and interpret the piece as one pleases, and the group opted to not perform on the piano or include balloons. The result, which seems to have predictably pissed off a bunch of people, is a moody exploration of destructive process, as an ambient backdrop of crackling flames is occasionally punctuated by percussive piano strings breaking or bits of the piano caving in. In the context of the rest of the album and clipping.'s heavy use of machine noise in their compositions, it suggests a strangely slow and quiet horror, a transformation from more conventional musicality into something more alien. I find the piece troubling, I don't quite know what to think of it, because the length (18 minutes), the lack of action, and the (to me) soothing sound of crackling fire seem to butt weirdly against the aggression of the act (setting a piano, possibly a perfectly usable piano, on fire).
I love this shit because it takes seriously Lockwood's composition but reframes it in performance and context, revealing that the procedure is never JUST the procedure, it's also all the stuff around it, all the tweaks to the procedure and what they mean, how the final product is framed and contextualized, how much of the process is captured and documented... there's so much you can talk about here. The means and the end are strangely comingled, and also demand, due to the often imperfect results, interpretive work from both the person running the procedure and from the audience experiencing the result. The strange Rube Goldberg nature of the procedure also radically changes the terms of art work: anyone, even without musical/artistic/performance training, can follow the instructions, but that doesn't make any particular performance meaningless or rote because everyone who runs the procedure can put her mark on it.
And yeah, anyone could do this art. So fucking what? Art increasingly functions less as an aesthetic experience and more as asset. Switzerland's Geneva Free Port, for example, hosts untold millions of artworks never seen by any human and indeed never moved at all, but sold frequently, on paper, on balance sheets, gold beneath Fort Knox. This art exists simply to exists, to sit there and represent a totally arbitrary value point agreed upon by elite capitalists for tax evasion purposes. Process, performance, and participation-based artists also find their work sucked into this nexus, but I think it puts up a bit more of a fight. At its best, procgen asserts that yeah, anyone CAN do art, when given the proper tools, so treating that art as a singular masterpiece that not coincidentally is a great store of tax evasion money makes no sense. The proper place of even the most skilled performances of this art is not a Swiss vault but... everywhere, enriching everyone's lives.
I bring all this up because once again, just like last time, the task fuckin falls to me, whether I want it or not, to actually try to situate cryptoart into an art historical context. I mean, is Christie's gonna do it? No. Are the galleries showing Cryptopunks gonna do it? Absolutely not. As I pointed out last time, and as we'll see today, even writing promoting NFT art shows little to no interest in talking about art, not when they can gush about limited supplies, cryptographic signatures, and investments investments INVESTMENTS! So somehow I'm left as the one person who can look at the breadth of over a century of avant garde developments in art, and recognize their culmination, their final apotheosis:
This motherfucker.
This Happy Tree Friends genetic reject is a "Cryptokitty". I want to talk about what a cryptokitty is, but because the backend is more interesting I gotta get the aesthetic critique out of the way first. The aesthetic is... fine, I guess? Worse than a neopet, better than a funko pop, is about where I'd rank it. They all pretty much look like this with a few ultra rare exceptions, and it's just not to my taste. The eyes are a bit too Western Adult Animated Comedy Series, and looking at many of them causes the chill apparition of a dead 9gag poster to manifest behind me and groan "lol derp". Chilling stuff. This is the kind of thing that if not for the price tag on some of them of over a MILLION GOD DAMN DOLLARS I would just write off as not for me and move on, like minions, or Among Us, or the male gender.
They look like Flash animation characters because they essentially are: built in Scalable Vector Graphics, the semi-functional format that utterly failed to act as a fallback to Flash after Google and Apple bilaterally murdered it. Their color palettes are limited to three or four tones, which are consistent throughout the body. What they lack in morphic differentiation--they almost all have an identical body template--they attempt to make up for in fur patterns, colors, facial expressions, and occasionally accessories like horns or a dracula costume.
Cryptokitties market themselves with this procedural--and gamified--element, making cryptokitties a good example of process-based cryptoart. The cats can be bred to get new color and style combinations. However, the rate at which they can be bred slows down over time, so you COULD treat cryptokitties as a game of optimizations. How can you best get the kitty with the qualities you want? You might also treat it as exploratory, discovering the serendipitous results of the procedure. Procgen text accompanies each kitty, and while it's a bit lolrandom and tonally inconsistent (it seems to follow a rudimentary grammar system without controls that cause later sentences to be affected by earlier ones), that's still something you might explore. Hell, I kind of enjoyed the few minutes I spent looking over enough bios to get a sense of how they operate, even if it seems about at the complexity level of fairly simple Tracery bots or below. And I have to admit, I found myself swayed towards liking them a bit better, if only because the illustrations have some command structures embedded within them that ensure things like: coloration on tails match bodies, accessories and scenery don't clip weirdly into bodies, patterns on the bodies often carry over to the tails depending on the tail type, &c. It's not super sophisticated, but there ARE some controls there, which in this space is a god damn miracle of design.
Here's the thing about the optimization strategies of most crypto art. It's boring. There's an aesthetic pleasure that tons of web games operate on, the pleasure of Number Go Up. Narrative betting sims like Blaseball uses it, resource management games like Fallen London use it, clicker games like Cookie Clicker use it, and most infamously of course gacha games of all sorts use it. But if all you want to do is hit a button to start a number going up, you don't need to buy an NFT, you just need to buy a stopwatch.
In some ideal fantasy version of reality, cryptokitties would be an answer to that: a gamified version of cryptoart that plays with the complex engagement of procgen: both appreciating kitties for their unique qualities, and appreciating what gets gleaned of the underlying mechanics and how different builds might be developed to get certain outcomes and so on. Both the result, and the rube goldberg machine used to get there. This version would open up to the masses the long tradition of inbreeding and induced deformity that rich freaks have sought in their pets for ages. Like instead of creating abominations saddled for life with disastrous health defects, simply for show, create a cryptokitty instead! After all, no animals actually get hurt.
Oh, right, ah, except of course for the whole "apocalyptic impact on the global climate" thing that's so intrinsic to cryptocurrency's economics of conspicuous waste. Turns out this cat inbreeding scheme just adds a few extra steps to the process of environmental degradation in the name of kitsch.
So maybe what I ACTUALLY want is something open ended, something more like a picrew with randomness. Picrew is a platform where cartoonists can provide a library of, essentially, paper dolls that people can build icons from. Imagine that but with more nuanced controls and an emphasis on random generation, maybe a visual programming language that could be used to avoid the collisions and aesthetic incoherence that is inherent in anything where you just roll the dice. An open engine for this kind of thing.
But that wouldn't be compatible with the real point of NFTs, of course, which is to make a few fucks at the top of the pyramid scheme a whole lot of money.
I'm afraid that all my grudging respect for cryptokitties dissolves when I remember that none of the generative strategies they use matter. That's the message I've gotten from the marketplace, at least, which is mostly incomprehensible to me, but seems above all to privilege, simply, whether a cryptokitty is first generation or later. You can pick up kitties from the 45th generation for only $16 and an eternal mark on your soul. This is by design: all the breeding stuff is bullshit worthless nonsense by the explicit design intentions of the devs, who set out to make gen 0 the most valuable by limiting forever the amount of kitties in that generation.
So sweep aside literally everything I've said about the APPARENT Rube Goldberg contraption that is cryptokitties and behold the real one: the same gold bug economics that all of cryptocurrency runs on, a deranged libertarian conviction that currencies that people use to buy basic shit for survival should ONLY EVER BECOME MORE RARE AND VALUABLE. In the first one of these I said I didn't want to talk about Blockchain cause it's stupid and boring. But I arrived at that after going through this frustrating emotional relationship that involved comprehending the incredibly clever structure blockchain uses to achieve a deflationary basis and trustless records of transactions, getting kind of impressed, remembering that this is a Rube Goldberg machine made of fucking corpses, getting insanely mad, then eventually just getting bored. They managed to invent a Rube Goldberg machine that on final examination is just banal: the ultimate banality of finance capitalism, which not even The Wolf of Wall Street could make look anything other than repulsive and tedious.
Understanding that this is all a huge dumb alchemy mechanism for turning coal into a virtual gold by way of Chinese power plants helpfully clarifies why cryptokitty design contradicts itself fundamentally. The breeding mechanics and extra bullshit? It's just a pretext. The Rube Goldberg really is a mechanism for generating token uniqueness: literally, unique tokens, non fungible tokens, tokens that hold different values due to their attached artistic content, and so cannot be used interchangeably. But also, token uniqueness, uniqueness that only exists on a surface level, because it only needs to be surface level in order to fulfill a minimum of it's pitch to investors.
Once you comprehend this it makes more sense why competitors to cryptokitties don't even try not to look like complete dogshit. Did I go a little easy on cryptokitties? Well, it's probably cause I researched the whole field before writing that individual section. I knew what was coming. Folks, take a look at Hashmasks:
Hashmasks come from corporation Suum Cuique Labs, a company founded by economics majors and based in--and I'm sure this is purely coincidence connected in no way to the stuff I said earlier about the art market--Switzerland. Their name is latin translating to "To each his own." Hm. They are randomly generated by slapping together a body type, a background, an eye color, and a mask, which I guess does count as a "process" in some sense, though not much of one. There are a few variations of each broad mask archetype, and background get recolored, but unlike cryptokitties there seem to be no controls that would allow one part of a hashmask's morphology to affect another part.
What stands out to me about hashmasks is that they're as ugly as they are lazy as they are appropriative. Look at the weird pose they're all in. Isn't that something? Isn't that fucking weird looking? It's cause a few of them are holding items and the developers were too lazy or too incompetent to put in any sort of controls that would have a torso style and arm position shift to accommodate those details. This is, and I cannot stress this enough, basic ass shit that any self respecting Picrew creator would not accept. Many of the sprites have weird transparent bodies which maybe is supposed to come across as artsy but mostly just looks like a generation error. The designs are both noisy in the individual (at least cryptokitties had color scheme generation controls!) and dull in aggregate, quickly revealing themselves as profoundly repetitive. There's no real generation strategy to speak of here, just greater rarity of certain individual features popping up. From this lunatic gold bug perspective, perhaps mass hideousness of the images are not a bug, but a feature, driving up the value of those pieces that don't inspire revulsion.
I feel like maybe the incredible contempt I have for these might offend some sensibilities. Didn't artists work hard to create the individual components of these awful trainwrecks? Yeah, I mean, maybe, but I'm not sure how my calling these out as ugly as hell is more offensive than the incredible contempt that Suum Cuique shows for its artists and its buyers. Look how the bottle in this image sorta awkwardly sits on top of the arm, with no attempt made to do the bare minimum of integrating the two component images:
The message of this artwork is "fuck you, hog, you'll take what we give you," and I actually think it would be more offensive to waste your time and mine by critiquing the component art here in good faith. The unique never to be repeated ever on The Immutable Blockchain name for this piece is simply "pepe". Let's not pretend anyone involved in actually profiting off of hashmasks as either corporate heads or cryptoart speculators is looking at these because they just love the backgrounds. If I like some of the pieces, the best I can do to honor them is to excoriate the artists' bosses.
Just who were the artists behind hashmasks though?
Oh I am so glad you asked. :)
The website brags that these icons were created by "over 70 artists" supposedly spamming the globe. That was autocorrect for "spanning" but fuck it I'm keeping it in, it fits well enough. This description might just be enough to head one major critique of hashmasks: that they are grotesquely appropriative, quite literally culture as costume. Suum Cuique helpfully provided a way of searching for particular features, such as mask types. One of the options there concisely just reads: "African."
Now, is it just me, or is there's something a bit galling about a company from famous Nazi gold funded economy Switzerland creating Supercopyright Art of nebulous tribal masks and Basquiat ripoffs? Oh hey as an unrelated side note, I looked up that "to each his own" slogan the company's named after and there's some VERY interesting historical uses listed under 'Motto'.
NEAT!
Anyway, now that I'm done going shrieking deranged from that brief bit of research, I think I was talking about them ripping off Basquiat?
Basquiat, street artist turned gallery darling and protege of Warhol, isn't the only person whose style and methods were ganked to make these, either. I spy color field painting reminiscent of Frankenthaler and Rothko, some noodly images that look suspiciously like knockoff Harring, and of course, in the example above, a random fucking chapter from Moby Dick. Sure! Based on all that I've said about procedural art, it might come as a surprise that I despise this. But there's such an obvious qualitative difference between the appropriation happening here--cynical, meaningless, suggesting less than nothing by its juxtapositions--and the reframing taking place in clipping.'s rendition of Piano Burning.
Style isn't copyrightable, and on the whole that's a good thing. The turn towards process is also a good thing; everyone should have the opportunity to create art. From that perspective, it's true that Suum Cuique Labs infringe nothing with these pieces. From an art perspective, though, that's sort of beside the point. They cynically take advantage of open culture to enclose it again, aspiring to use blockchain to create Supercopyright with ludicrously high buy-in. And more to the point perhaps, this free ganking is a tacit acknowledgement that the artists they got were simply not good enough to achieve anything other than a bad copy of a dead black artist's style.
But ah I never did get around to talking about those 70 global artists did I? That was an oversight on my part! Let me just check on-
Oh, that's odd, it's impossible to find any CREDITS for any of this art? That information is completely absent from the site? And the company founders are entirely anonymous, going by "John Doe!" How curious that is. So I can't really talk about the artists at all, can I "John"? Not that I'm alleging that they don't exist or anything like that, of course! It's entirely possible that this really is a global art group and not just a slightly more talented Hans from Stuttgart producing Maori mask pastiches for South African apartheid profiteers like Elon Musk to speculate on.
But if that's the case why are the artists uncredited? Why not put them front and center?
What is front and center is a whole lot of information about these tokens as a store of value and generator of wealth. Oh and some stuff about how you can name and group these hideous things if you choose. Actually that's one of the selling points. The website proudly displays people's collections of these things with people's attempts at goofy curatorial statements. The consumer is an artist! In fact, in the absence of any artist credits, and in the complete alienation from the ability to produce art personally, the consumer is the ONLY artist. Buying the right products to put on a virtual shelf is the only artistry in this space. Pretty bleak stuff.
Oh, it's worth noting though that I can identify one person involved with the project. Their community manager outreach person or whatever actually got in touch after my last article to explain that yes he had definitely read the piece and also the future of NFTs was bright! After a predictably imbecilic conversation in which he resolutely failed to understand that if I have a jpeg on my computer, I've got it, I have the image, I own it, basic stuff that I would've assumed was incontrovertible obvious fact, I prodded him into revealing he HADN'T read the article, and he flounced angrily out of the conversation. It's possible he'll actually read this one and get mad, but well, if he was interested in putting in any effort, he wouldn't be involved in such a slipshod operate as hashmasks to begin with, would he?
There's one final semi-procedural NFT art nonsense I want to touch on, but just as cryptocurrency in general sent me through stages of fascination to rage to boredom, I find myself at the end of this article, burnt out and bored with "cryptopunks". Like, here's the thing, right:
I, too, read Diesel Sweeties in 2004.
Nothing about cryptopunks looks visually particularly impressive to me. It seems from news coverage like maybe they were procedurally generated, but if they were they seem to have been pre-baked and then released after curation. Whatever. Like I said before, this stuff is token generation. The point is just to churn out a bunch of images as lazily as possible to justify their use as a speculative asset.
Hashmasks deserved analysis simply for being so offensively cynical. Cryptopunks are so bland that I just can't even work up a good head of steam.
No, I'm mostly interested in them because of their place in the art world. Cryptopunks have appeared in at least one actual gallery: the Kate Vass gallery- sorry, "Galerie" included them in a show a few years ago with other blockchain artists. The enthusiastic writeup on the cryptopunks website, however, doesn't talk about what it means to be in an art show with artists like, most notably, Ai Weiwei. There's no situating of the art in a context, no comparison with works that are also in the show...
No, what they spend time on is, you guessed it, the technology they used to verify that the prints were in fact one of a kind. If you wanted to SELL the print you would have to unseal a wax-sealed envelope on the back and sell the associated NFT on the blockchain as well, or, something. This all sounds pretty stupid to me, but it's what they decided to present as their great artistic innovation in lieu of any other critical framework, so, fine.
The gallerie is no better. The page for the show announces three major talks for the show:
10 YEARS BITCOIN - A BIG IDEA CELEBRATES ITS BIRTHDAY (not about art)
THE CRYPTO VALLEY ECOSYSTEM (not about art)
ASSET-BACKED COINS: ART COIN & CO. (not about art but art tokenization)
It's clear that the gallery also is disinterested in questions of art history, meaning, theme, aesthetic, affect, social impact, or any of the other countless values that art has held at one time or another for critics through the centuries. No, they are interested in asset management, and in particular promoting cryptocurrency as an investment vehicle. One would think that a galerie would show more interest in the actual artwork on its walls! Strange, I wonder what this place-
"Founded in 2016 in 🚨 Switzerland 🚨"
[slams fist on desk] THEY CAN'T KEEP GETTING AWAY WITH IT!!!
I can't say I'm overly impressed with the gaeleirie's overall presentation of its shows generally. It focuses on contemporary digital art, and unfortunately tends to present this art not in a wider history of abstraction but instead breathlessly focuses on "algorithms" and so on. But nowhere is it so disinterested in the content of art as when it touches on crypto. The message is clear; we're all here for one thing and one thing only: speculating on tulips. The color of the bloom is completely beside the point.
I think it's inevitable that cryptokitties will be a strange outlier in this field. Its contradictions between wanting a system and ultimately only needing a pretext for financial speculation means that its inevitable competitors will be far more boring and lazy. This process art looks much less like a Rube Goldberg contraption, beautiful for its process and the way the individual artist/performer enacts it, and a lot more like factory production again. They've just hit on a way to make the deep contradictions of this process work to their benefit: if they're forced by the market to embrace automation to produce their product, they can turn that automation and its completely arbitrary limited use into a point of prestige.
Cryptocurrency, as a pure expression of libertarian brain worms, portends where capital as a whole is headed, and the lazy implementations we can expect in the future from procedural art. I do think we're rapidly reaching a point where elements of artistic production can be outsourced to more sophisticated neural net systems. Oh, I don't mean that we've reached a point where the machines can just make art. No, what they can do is smaller menial tasks. Generating large lists of things that are plugged into a manually curated procgen grammar system, for example. I've done this before for my procgen games! It's very handy, when I'm feeling a bit creatively stuck, to take an existing grammar and dump it into a site like Text Synth in order to see the surreal outputs the machine suggests, then incorporate them into my own games. Elsewhere in the industry, voice actors worry that their recordings will be used to generate AI models that can fill in "barks", short utterances in games that are often repeated, contextual, and throwaway. The technology to rapidly alter and finish artwork continues to expand, and while a lot of NFT art has the look of "existing image with a bad photoshop filter slapped on", the G'MIC library and its implementation in libre art program Krita is essential to my workflow at this point.
When I wrote last time that we're seeing a deskilling of artistic labor, some people got a bit upset at me. How could I say that digital art was unskilled, which is to say, not good! Well, I didn't say that, in point of fact, and in fact said the opposite, but I stand by the claim that digital art is increasingly easy to produce and that means that the individual average laborer in this job market will be able to command less value. You hear about those Luddites? Guys just hated technology! They smashed up all sorts of, ah, what was it again? Oh right, looms that were transforming their previously skilled cottage labor into something that could be done by anyone employed by a factory. Now, those looms themselves weren't evil! In the right hands and with an equitable model of running an economy, they COULD have meant an explosion of creativity and a higher standard of living for the masses. To an extent they even did do that, just at the cost of emisserating a previously skilled class of laborer and enriching a bunch of factory owning cunts.
Various avant gardes of the last century and a half took a long hard look at the implications of automation for their fields and decided to pick a different path than the Luddites. They embraced the possibilities of procedural generation and proclaimed that anyone could, in the end, also make art! Anyone can, after all, follow the procedure to produce a performance of Piano Burning. Anyone can take a rubbing from a wood floor and derive creatures from the result. This is potentially incredibly empowering, as is the beauty of the process and a focus on creating interesting Rube Goldberg machines for spreading art to the masses.
NFTs have less than nothing to do with that empowerment. On the contrary, they restrict production back again to people with the familiarity with the ugly, restrictive Rube Goldberg contraption of the blockchain and the ability--often through established clout--to command interest from people with Swiss bank accounts. They race to the bottom, closing off all the most exciting aspects of procedural art in order to generate token uniqueness with the greatest of slapdash efficiency and least actual artistic labor overhead. Eventually, even the technology used to generate the NFTs becomes obscure, a footnote, barely understood by breathless tech and finance reporters, handwaved by art galleries more interested in the art as a compact tax haven.
This isn't an indictment of procedural art and the beauty of the Rube Goldberg contraptions we use to make it. It's an indictment of the free market fanaticism that underpins this art, of the small minds that conceived of NFTs, and, if we leave this revolution in art unfinished, if we let them enclose and arbitrarily restrain our creativity to an endless series of Limited Time Offers, an indictment of all of us who might be artists: which is to say, all of us.
No comments:
Post a Comment