[Content Warning: Highly Indecent Erotic Art]
That’s one of the third rails of philosophy. It’s like starting a land war in Asia, when you’re not the Asian.
I’m going to avoid going down that rathole, because it’ll just make us all miserable, or miserable-ish, though you can probably guess my position: “art” is a vague concept. [stderr] I’m aware that the “it’s a vague concept” position may seem remarkably like a dodge but I am, seriously, making that argument in good faith: it may be impractical to define “art.” Especially since someone attempting to define “art” may be carefully constructing their definition in order to privilege humanity – as often happens with “creativity” and “intelligence.” It ought to be enough to point and observe that there has been a lot of argument about those concepts, and where there’s smoke and confusion, there is not clarity; simple definitions of “art” do not appear to exist.
Pan and the goat, found at Herculanaeum in 1752, dated tp 1st century AD [source]
My attitude about art is strongly influenced by my childhood experiences. My parents both had an extreme love of medieval art; I think I have been in most of the great museums of the US and Europe, and they all have a section devoted to medieval art. As a child I quickly developed the opinion that medieval art was art from before when humanity invented good art. With many many hours in The Louvre and Les Invalides, I became very fond of Napoleonic art, Gerome, Ingres, David, Gericault, Detaille, etc., But after a while I concluded that what they were doing was what happened before photography had been invented. I did come to hate medieval art (most of it) and would argue that art was not invented until the renaissance. At which point my father and I would melt down into a hand-waving match which usually ended with him taking the position that there was a great deal of wonderful Greek and Roman art, and I needed to fill in my ignorance and me taking the position that “that doesn’t make medieval art suck any less.” I mean, seriously, I would say, “I can compare my refrigerator drawings with the Bayeux tapestry and the main difference is that admittedly tapestry is a more difficult medium, but geeze that stuff looks like it was pre-figuring cubism.” Or something like that.
Frontpiece of the church of St Foy de Conques, Conques France, 9th century
Anyhow, I can’t understand why people are impressed by the mass of medieval art, almost all of which has something to do with christian absurdity or the power fantasies of monarchs. The church at Conques was a regular summer visiting-spot and I came to hate the art. In fact, (and this is the point of this posting) you’d be hard-pressed to get Midjourney or some other AI to do such a crap job of designing a sculpture for a church. Look at the proportions of the hands and feet and heads. These are all mutant gnome people or something. Obviously, they had noses at one point, so I won’t take off points there but this is not Rodin-quality or even ancient Herculanaeum-quality stuff.
This is one reason why I am not particularly sensitive to AI’s tendency to play fast and loose with finger counts, etc. Normally I would have put some effort into shape-correcting this quick shot I took with my camera phone, but really why bother?
late renaissance art, Carnegie Museum, Pittsburgh
I was wandering around the Carnegie with Anna, and I kept noticing horrible hands and feet and distorted proportions that would have totally embarrassed an AI. But this was great art in a famous museum. So things like weird distortions don’t count, right? It’s still great art. Right?
Allesandro Allori, portrait of Isabella de Medici, Ca 1570
Sure, there are some problems with the paint not sticking to the board, and apparently the museum did not use anti-reflection coated glass, because there are reflections all over the place, but – seriously – can you imagine the gales of uproarious laughter if Midjourney burped out something that bad? For one thing, it would have possibly rendered her as a hottie anime-style waifu. And now I guess I have to see what that would look like, right?
While it’s doing that, here’s something that beats the snot out of medieval art:
Midjourney AI and mjr: “a frieze of a medieval church representing cthulhu and his apostles, carved in shiny black granite for the ages, ia! ia! Ia!”
And Waifu De Medici:
Midjourney AI and mjr: “waifu style version of isabella de medici [image prompt]”
I like how the AI gave up on the details of the cupid’s face and just kind of punted – the painter knew that everyone was going to be staring at that completely exposed pink ankle with the correct number of toesie-woesies. But look at her hand. Ow, she should keep her hands away from the tool-rest when she’s working with a lathe, that is a nasty fracture.
I do not specifically hate this great art. It’s in a famous museum, so it must be great. In fact, the mere fact that it is in a famous museum makes it, ipso facto, prima facie, wolens nolens great art even though I am pretty sure Stable Diffusion could completely dust that if I cared to bother firing up ComfyUi. Of course it might be a group of goth punk waifus on motorcycles, for all I know.
Some of you are no doubt thinking “but there’s no creativity in that!” in spite of not knowing what creativity is and having a definition for it that does not depend on the idea that creativity is only something humans or human proximates can do, or something like that. But here’s the question: if you were asked “do a portrait of so-and-so and her handmaiden, mostly nude, and definitely showing off her ankle” would there be some magic additional creativity that gets thrown at the problem? That is, frankly, a pretty boring image as long as we understand it as art out of its historical context – it’s not very elegant erotica at all. I mean, it’s way better than the Bayeux Tapestry version would be, but it’s not as good as what Caravaggio would have done, and even Caravaggio, if paid to produce an un-creative wall hanger of some lordling’s girlfriend – would have done something better than this but there are limits to even the creativity of a great artist. As soon as we start thinking like that, we have to divide Caravaggio’s output into “great art” (whatever that is) and “hackwork that the Pope paid for” and how can we tell the difference?
There has been a lot of swirling thought around the question of “AI art” and whether it’s art and whether AI can be creative or not, and whether or not they can create great art – but if you look at some of the human-generated great art I present above can you seriously tell me anyone should care? I see a shortage of greatness on the human side. Again, I will leave you with the thought that it’s not the Caravaggios who are threatened by the AI, yet, but rather the hackwork artists who will no longer get hired by the pope to produce virgin Mary spankporn, because they’ll just get Stable Diffusion to crank out as much of that stuff as they can look at before their eyeballs dry up.
But do you have to be creative to make art?
The best I can get for a definition: art is anything that some humans consider art. Which is obviously circular and relative
But in my defense this position is performance art
i remember getting this question in my “art appreciation” class – my technical art school’s one requisite token nod to the intellectual end of things. i was an abjectly commonplace anti-intellectual asswipe of a white boy, like to think i’m better now. my marcel duchamp post had a feint in this direction. the main thrust of the entire 20th century of art was setting people free of the chains of “shit must be done like X or is no bueno” and the anti-ai mob is reeeeally quick to throw that entire era of thought and creation out the window.
man that was a poorly composed paragraph. i was extremely distracted.
I don’t know that the Bayeux Tapestry was made as great art. It’s as much a political chronicle and reflection of the methods of establishing the divine rights of the Normans over Britain.
It’s propaganda meant to hang in a church to display wealth and power. It isn’t particularly finely crafted, but it’s a huge flex in terms of the size of base fabric, and the amount of work it required to embroider all that yardage. IIRC the borders are woven into the tapestry, so there must have been an entire convent behind its decoration.
One thing that you always have to keep in mind is that just because the figures in the Baxeux tapestry or Egyptian reliefs look weird or unproportional, doesn’t mean that the people designing them were hacks, anymore than Mickey Mouse having giant hands and only four fingers means that cartoon artists can’t draw proper proportions.
The thing about Medieval art is that you really need a Medieval mindset to appreciate it. The fact we instinctively go to it with the critical framework we do is testament to how thoroughly Renaissance-derived cultural propaganda has seeped into our world view. The whole idea of “great art” and the “great artist” is a Renaissance invention, a trope to heroise the cultural producers of the Northern Italian city-states, whose military might was rather shaky in the face of the Germanic (hence, “Gothic” and “Barbarous”) powers of the Holy Roman Empire. Renaissance art was all about the rejection of the Medieval and the embrace of a constructed vision of Classical art in opposition, because that culture was one key way Italians could raise themselves above their North European rivals. Our appreciation of “good art” is still heavily influenced by Renaissance Classicism. And, with it, a model of cultural improvement and advancement – obviously Classical Greek art was an advance over Archaic Greek art, obviously Renaissance art was an advance over Medieval, just as we moderns are obviously more advanced and better people than our predecessors in terms of culture and morals and politics. That concept, most explicitly formulated by WInckelmann in the 18th Century, is the very heart of artistic “modernism” that 20th Century thinkers were trying to challenge.
Medieval sculptors and painters and weavers were not short of Classical models they could emulate. Likewise with Medieval writers, who could (and some did) have copied the style of Vergil, Ovid and Cicero. The fact they didn’t shows they had other concerns, other aesthetics and other appreciations of the world. Photorealism, mathematical proportion, perspective – none of these things really mattered to Medieval artists. Indeed, the idea of the “artist” was something of a foreign one, given that the production of visual culture was seen a artisan’s trade, not a high intellectual calling. The “celebrity artist” was not something the Middle Ages in Europe really had, and the idea you would sign your work, or have your name attached to it, was almost unknown. Indeed, very few earlier cultures had this idea either. There were a few celebrated Greek painters and sculptors like Praxiteles and Apelles, but the majority of ancient painters and sculptors were as unknown to their contemporaries as they are to us.
Why do we still value proportion and realism and the evocation of particular symbols and stories and ideas? It is just as irrational to value these things in art as to value what Medieval viewers and creators did. Obviously we do, because that’s what our culture has led us to value, and we tend to see it as unacceptable chauvinism when we hold up our modern European artistic sensibilities as better than the art of, say, Africa or Asia. But when it’s our own historical art in the spotlight, suddenly the inherited model of Classicising progress that derives from exactly the same place as the colonial dismissal of other cultures’ art is in full force and implicitly unchallenged.
If, one day, we ever do build a machine which can effectively communicate and someone criticises its ability to draw hands I believe it will inevitably respond with some version of “I learned it from YOU!”
The common element here is mediocrity. The little sculptures on the church facade are thoroughly mediocre representations of scenes from the Bible or Saints’ lives or whatever, and the stuff Midjourney et al churn out are thoroughly mediocre representations of whatever the prompter asked for.
I’ve been thinking a lot recently about what exactly genAI systems do and how to explain it to people (there are people at work who want to encrapify things with GPT and I want to dissuade them, and precise arguments help), and I think the key thing to understand is that the output, while syntactically impressive, is mediocre. Not shit or garbage, and not excellent or high-quality, just mediocre. It’s impressive that an automated system can make something that good, but it isn’t really good enough to use for anything useful.
Thus, the only use cases are those where mediocrity is sufficient, and that leaves the ones we have already seen deployed: spam, scams, and disinformation.
Anyway: just because I can’t define «rubbish art» doesn’t mean I can’t recognize it.
cartomancer @ # 6: Photorealism, mathematical proportion, perspective – none of these things really mattered to Medieval artists.
More recently, we have experienced something of the opposite, starting in the 1950s as western Cold War strategists mounted an offensive against “socialist realism”. The promotion of “abstract art” (though building on the work of sincere individual creatives) came from a political agenda deliberately fostered through academic budgets and foundation funding, imposed on an unwilling bourgeoisie – see, e.g., Hugh Wilford’s The Mighty Wurlitzer: How the CIA Played America.
As far as “what is art?”, I had enough of an answer fairly early. I was perhaps seven or eight when I discovered M.C. Escher. Being the kid I was, I naturally went to find a book about art to find out more about the guy. And in the books about art that I looked in, I learned that whatever it was Escher was doing, it wasn’t “art”, because I’d look where he was supposed to be and find Max Ernst, every single time. Now, there’s nothing wrong with Max Ernst, it’s just that in my mind he is forever “that guy in the art encyclopedia next to where Escher isn’t”. By common consensus, the world had apparently decided that what Escher was doing was draughtsmanship or something, and not worthy of the a-word. And then I’d look at something by Rothko or Pollock or whatever, and have to process that apparently that IS art, but the intricate and fascinating impossible worlds Escher conjured weren’t. And eight year old me though the equivalent of “well fuck that”, decided I knew what I liked, and went into engineering.
What do you call a guy nailed to a cross on the wall? Art.
The definition of art will undoubtedly always elude us, it’s like trying to nail jello to the wall. Any attempt to be specific about what defines art is will exclude things which other people accept as art. Try putting “art” in place of “property” in Proudhon’s famous tri-lemma:
Art is Theft
Art is Freedom
Art is Impossible
I think that the discussions on this blog have argued for all three of these options, and others.
What Cartomancer wrote above, #6, is true. The underlying message I take from Cartomancer’s comment is that a lot of art has cultural aspects, some of which are no longer common today. That doesn’t mean that ancient art cannot be appreciated as art, but it’s likely that what we take away from that art will be different. Which leads me to my point, which is that art really falls into semiotics. Which is the study of things which intentionally or unintentionally communicate meanings or feelings to the viewer.
With a definition like that, one viewer could view an object, like a Kandinsky, and not have any meaning or feeling communicated to them. Thus, they could say that object is not art. But a different viewer could gather a great deal of meaning from the same object. Note that everything lies on the viewer, the object can communicate intentionally or unintentionally.
We often place the burden of defining art onto the creator. Is what the creator doing making art? Does the creator think they are creating art? Are the tools the creator using the tools of art? Is the creator’s mind different, and allowing them to create art?
Maybe that’s the wrong way to look at art? Maybe we should define art as something which conveys a message to the viewer.
Or is that view too expansive? Because that allows natural processes to be art. A sunset, a child’s smile, and a lynching would all be art whether it is captured by a photographer, or imagined by a painter, or generated by an AI.
Or is this view too restrictive? It allows people to say, “I know it when I see it, and that’s not art.” And people can say that about statuary in medieval cathedrals as easily as at brutalist architecture.
At the end of the day, AI will be a tool to generate art. It will take away some of the jobs of mediocre artists, probably the best paying ones for the bland corporate office market. But the artist who creates works out of enjoyment will survive. And artists who don’t need to create art to earn a living will survive. Even today a lot of artists are of that type, and just as there is still some demand for buggy-whips, there will always be a market for even mediocre human-created art as well as exceptional human-created art.
@6 cartomancer – Neil Gaiman said that the Renaissance was basically a bunch of Italians poncing around in fine clothes and saying they were the reincarnation of Classical Greece and/or Rome.
But seriously, the rediscovery of the ancient culture (Greece and Rome, natch) did have a lot to do with it. It was a continuous process, indeed starting in the Middle Ages and at last reaching critical mass, exploding in what we now call the Renaissance.
That said, there has been no culture which has not imagined itself to be ground-braking, unique, unprecedented and sooo damned cool.
And @8 Ketil nails it, mediocrity is the word. But as a partial defense, artists need not only quality but also volume of production, so sometimes they gotta knock off a quick sketch/oratorio/terzina for a quick buck, ’cause one needs to eat.
Further, even when crafting something wonderful, they are not conscious of it and cannot describe or recreate the process – everything is so damn indeterminate, which is a typical attribute of art I suppose.
They get the proportions wrong and they are mediocre hacks. Rodin gets it wrong and he’s an effing genius.
The Stubborn Genius of Auguste Rodin
Maybe, maybe not. If I am raising my my little fingers as if to grasp a tea-cup pretentiously, my little fingers and only my little fingers, reach a point where they ‘click’ back and up into just that position relative to the rest of my fingers.
As for the great art business, I am suspicious of the whole idea, having first encountered the idea as a way of enforcing a certain kind of intellectual superiority, one that I was clearly not up to achieving because of the frailty of my inferior female brain. I tend more to the William Morris view “‘Have nothing in your houses that you do not know to be beautiful or believe to be useful’, but I amend it to “Have nothing in your house that you do not believe to be beautiful or useful.”. I am suspicious of having anyone else pass judgment on another’s taste as in the UK it is so often an indicator of class and used to denigrate the masses, and their awful plebian tastes, more discreetly than when I was young, but that vicious sentiment is still there if you look. Why for instance does opera in the UK require not only the sort of ticket prices only the biggest of rock stars can command, but a subsidy from the public purse to support it in addition?
People hanging their hats on AI art being mediocre got a rude awakening in store. Or can keep their heads in the dirt forever, that’s certainly a way you can go. I ain’t saying all AI art is brilliant, but to say with certainty that it can’t get there seems willfully ignorant. Some is already there *right now* if you bother to look for it – or better yet, if you bother to make it. Go onnn… u kno u wanna….
Incidentally, I am a lot more liberal than Marcus about what I consider good art. I think during the medieval period there was probably some kind of brilliant art happening somewhere, maybe hiding in marginalia or in a craft that we aren’t looking at here. Maybe it was lost. There’s this way of drawing that came out of the late medieval period and carried into the less fancy stuff of the renaissance era that I’m rather fond of. Little dolly guys getting chopped up impassively on fields with tiny castles and bizarre animals. Good times.
I’ve got little to say regarding the qualities of ‘great art’, AI produced or otherwise.
Greatness seems to be a term bestowed by others, often quite remotely positioned from the creation of any particular piece of art. The decision that this one thing is ‘great’, and this other thing is ‘ephemera’ or ‘trash’ is a value judgement made by the person choosing – and it speaks as much about the chooser as it does about the thing being discussed. Witness the clamouring of the ‘white statues avatar’ folks aghast at the idea that their virginally pure greek marbles might once have been covered with gaudy pigments. Is David greater in stark white, or would he be better in vivid pink with glowing cheeks?
If art isn’t considered ‘great’ when it’s produced, can it subsequently become ‘great’ by judgement (whose?) of later observers? Does that mean that it was always ‘great’, or that standards of ‘greatness’ change with the needs and desires of each instance of cultural evolution?
I don’t really have a dog in that fight.
But AI? Seems a played-out force, to be honest. A tool for quickly and cheaply generating with a computer kinds of visual representations that could only previously have been made by paying an artist a minimal sum to quickly and thoughtlessly paint/draw something. I say ‘quickly’ and ‘cheaply’ here – but we really have no idea what the costs are/will be. Uber used to be cheap too, when they were still hemorrhaging VC money to try to create the monopoly required for profit-taking. AI is still in that golden, VC-subsidized stage of its existence.
I will say that the rate of improvement seems to be slowing drastically. All possible sources of cheaply available human-produced-and-categorized art that could be ingested by the models have been – and now the internet is being quickly poisoned by AI-generated art that will cause irreversible degeneracy in the models. The only solution that could allow further growth is ever-more-intense processing, but that has its limits as well, thermodynamic ones and societal ones. Unless Sam Altman gets his wish and we develop vast clean new sources of free electricity to run his models, we might be nearing the asymptotic peak of ‘AI’ performance from the current available technologies.
Sure, you can find lots of people feverishly prophesying the AI apotheosis, but since many of those people were just a few years ago hawking Blockchain as the ultimate solution to the world’s problems, I have a hard time taking it seriously. The current AI generative systems are likely to remain around as cool tools – a program that can generate images based on keywords referencing things that occurred before 2021, for example, or a program that can generate text that reads like the output of a corporate marketing committee that took a bunch of ketamine on a retreat, or a program that can generate video with weird artifacts that remind me of taking threshold doses of DMT, all of this is genuinely cool technology. Less globally useful that something like hypertext or bittorrent, but way cooler than that image ‘morphing’ that was briefly the rage in the late 90s, so there’s that.
World-changing? Hardly. Great art? That’s going to be for others to decide for themselves.
@14, Jazzlet wrote,
I think you and William Morris would be in violent agreement.
Because the sense I get from his writing is exactly what you wrote. Your version is a little clearer in modern English usage, but he is saying the same. As Marcus writes above, to believe something is “blue” is also to know something is “blue”. Because, in this case (not necessarily in all cases), the knowledge of “blueness” is undefinable, covers a wide range of phenomena, can be very personal, and is necessarily vague. To say you know something is “blue” is equivalent to saying you believe something is “blue”.
The fact that Morris propounded a certain aesthetic that he liked, I would attribute to either a) him sharing his own sense of what is beautiful, or b) being asked by people who don’t want to think for themselves and want an “expert” to tell them what is beautiful. There are a lot of human beings who fall into the second category, entire herds of interior designers depend on this population. It’s like telling people that they are all individuals.
Marcus
The art that survives from the early dark ages period into the 1000s is pretty spectacular. The octagon chapel built by Charlemagne at Aix/Aachen is gorgeous. The bronze wolf doors are easily on par with the goat in terms of sculptural quality. Sutton Hoo and Oseberg are 700s and contain many amazing pieces of artwork. The Book of Kells is yet another example of truly stunning art from the dark age period.
Mid to late Medieval art is highly controlled by the church and its ongoing battle with graven images. The Black Death was not a great period for European art.
I think the mere fact that St. Foy has been there for a millennium is rather impressive. However, that lecherous old goat is carved from fine marble and appears to be primarily decorative sculpture. The facade of St. Foy is multiple blocks of primarily structural limestone that are carved in bas relief into a frieze. It’s simply not possible to carve limestone like marble.
As a way station on a Pilgrimage route, its main purpose is as a medieval highway sign. ( food, shelter, relics!) The feet aren’t very visible from the ground perspective, but the fact that the angel on the right breaks the frame is very impressive stone carving. It’s obvious that it’s missing its paint, as is the interior. Medieval interiors were brightly painted in wild polychrome. Of course they look dull and dark with just bare stone.
When it is fully restored it is highly esthetically pleasing, but the art of fresco with earth pigments is a highly specialized and therefore expensive undertaking. I did laugh at Nil metuentes. Fear nothing.
The ‘restoration’ of Isabella de Medici is IMO, bad.
The top of her hand is very poorly done, and her hairline is a travesty. There are other portraits of her by the same artist that show an oval face with a slight widows peak. Your AI version bears a striking resemblance to the wedding portrait of Lucretia de Medici, who looked very similar to Isabelle.
From SMBC 8-)
ART A Friend of mine in Tulsa, Okla., when I was about eleven years old. I’d be interested to hear from him. There are so many pseudos around taking his name in vain.
From The Hipcrime Vocab, Chad C. Mulligan, Stand On Zanzibar by John Brunner