Art is Dead, Long Live Art
AI "art" is not art
As anyone who has spent more than a minute online in the last year will know human artists have an AI competitor. And not just one but dozens, with more popping up every week. All promise sleek images instantly in any style you wish of anything you want at the simple click of a button. The cost? A small licensing fee and a prompt. A small, or large, chunk of text that the AI, be it Midjourney, Stable Diffusion, DALL-E 2, or any other, uses to spit out a vast selection of images. Images that are new, relevant, and impressive in a matter of seconds.
Sure there's some weirdness to the creations some times, mostly around hands, eyes, and distinctions between two distinct things (such as the branch of a tree vanishing but leaves still present in the top left of the image below) but the image generators have learned fast and now there are people falling for AI images being real images. Such as this “house in Bucharest”:
Is this a fad? A short term obsession that will fade away like NFTs? Doesn't look like it. In other areas of AI, voice actors are being asked to sign away rights for their voices to be sampled by AI and Netflix used AI to draw the background to a recent anime. Financial advisors are being replaced and Wall Street is using AI for stock trading. HR departments have been using AI for years. AI is coming for us all.
Artists, once so safe in our bubble proclaiming AI will never replace us, are now staring down the barrel of automation. So? Artists aren't going to stop creating art just because some jumped up Excel formula can spit out a few
ladies from time to time,
form intriguing concept art,
create a magazine cover, or win art competitions. Artists bleed and sweat for their work, that's worth something, right? (I happen to think so) but “Art is Dead”, apparently. With Artstation artists going on 'strike' and panic ripping through the art industries you'd think AI was an existential threat. (It isn't.) What's all the fuss about? It isn't even art!
Cue the responses “You sure about that?” thus epitomising the confusion between art and beauty. Just because something is beautiful does not make it art and art does not have to be beautiful.
A sunset is beautiful but it is not art, though a Romanticist painting of a sunset is art. Why? Art is the sublime, the transcendent, it is ritual, communicative, and demonstrative of a human impulse that says “Look! Look at what I feel, see, and think!” Art is from a specific perspective. Art drips with meaning and intentionality created consciously and sub-consciously by the artist. The sub-conscious is not to be seen as less than the conscious for it is the patterns being recognised, ideas formulated and expressed, problems solved, without ever having to actively think about it. It simply flows into the art as much as the conscious is laid down by the artist.
The AI does none of that. AI has no perspective, no intentionality, no deep need to express something.
Artistic creation, says Aristotle, springs from the formative impulse and the craving for emotional expression. . . the aim of art is to represent not the outward appearance of things, but their inward significance; for this, and not the external mannerism and detail, is their reality. (Durant, The Story of Philosophy, 1926)
There is a communicative element (and much more) to art that is not found in the material. Count all the brushstrokes of J.M.W Turner's Fishermen at Sea, count all the colours, measure the canvas, and x-ray it to see the layers of paint, yet you will not find the fragility it elicits or the experience of the viewer held within. You will not discover why the moon is placed where it is nor the boat nor why the painting curves inward at the edges and corners. You will not find Turner's perspective, emotion, awe, or anything else of that nature. All of that is between artist and viewer and the painting is greater than the sum of its parts.
The AI does none of what the artist does and the viewers experience is lessened for it. When looking at AI “art” the viewer engages in pretence, wittingly or not, seeing meaning where there is none, feeling emotions that have not been intended or expected. The AI lacks all intentionality and the prompter has little control over the outcome. The prompter has also been conned. Paragraphs of text-to-image prompt may appear to be intentionality but you can only suggest, not control. Like standing behind an illustrator and saying “Why not draw the tree there, why not have him smile instead of frown.” The prompter, like a manager, feels in control when they aren't and when he learns of his lack of control sets down more paper work, more words, more definitions, that do nothing to advance his control but make him feel better for a time. This is the AI prompter. Tweaking wordage in the hopes his imagination can be realised on screen through an algorithm he doesn't understand. Even the programmers don’t know how the AI works. As the manager would be better served doing rather than outsourcing, so would the prompter. Learn how to draw, learn how to paint, put your ideas onto canvas with a brush that you control and understand. Learn colour theory, perspective, and everything else. Your creation doesn't have to rival Turner but it does have to be from you.
Will this take time? Yes. Will it take you away from Netflix, or the pub, or some sportsball game? Yes. Will it be worthwhile? Yes. But herein lies the selling point of Midjourney and it's ilk.
Speed.
Midjourney founder, David Holz, says exactly that:
Artists on the platform tell us it allows them to be more creative and explorative in the beginning, coming up with a lot of ideas in a short amount of time.
Right now, our professional users are using the platform for concepting. The hardest part of [a commercial art project] is often at the beginning, when the stakeholder doesn’t know what they want and has to see some ideas to react to. Midjourney can help people converge on the idea they want much more quickly, because iterating on those concepts is very laborious.
Yes it is “laborious” but to what end does it need to go quicker? Simply to keep up with the torrent of shit being served up for “consumers” (a grotesque word). How long before AI changes from 'a handy concept generator' into 'doing all the work' and the artists skill and imagination atrophies from lack of use? Given AI has won art awards, been used for book covers, and made a video game. I'd say not very long. These systems are learning fast and the artist using them to “speed up their workflow” may soon find it becomes their entire workflow. Artists reduced to machine prompters. How depressing. Humans reduced to robots; slaves.
With each technological advancement more of humanity is stripped away and what exactly is the point of stripping away art? So you can lie on the sofa 'consuming' AI produced anime while stuffing your face with human kibble and drinking weird sugar water. I guess Wall-E was the accurate prediction then. How depressing. No agency, no intentionality. A life of vapid consumption that will be forgotten within hours, minutes even, of it passing your eyes or lips. All subject to a machine-god, and bereft of will.
The person trained by the AI engine is nothing more than an AI prompter. Thinking like an AI and disconnected from their human nature. The AI is a robot, robots are slaves, why are you submitting your humanity over to a machine-slave? Let it do the tasks man doesn't want to do, not the tasks that give man meaning and purpose. Otherwise you will find yourself sweeping the floor and dusting the fans of your AI baron as it spits out portrait after portrait of its AI mistress.
But it gets worse. You do not own the tools with AI image generators. You do not control how it is influenced. You do not decide what art is fed into the machine, what quality of work, what style. You choose nothing. Your agency is limited, negated, ground into dust by some techie out in San Francisco who doesn't know you or care about you. Gone are the styles specific to a time, place, and people. Gone will the ebb and flow of art movements. All will become a vast homogenous infinite style controlled by nameless men tapping code under fluorescent tubes in a world that's forgotten how to paint. While there are thousands of style guides how do you tell the machine to generate something that hasn't been done before? What is your reference word for a new movement, a new aesthetic, a new vision of the world?
I'm sure someone is going to type up a bunch of prompts or find examples of Midjourney creating different styles but it is fake. An art movement is an organic thing, it's a group of disgruntled artists rising against the establishment, it’s old masters demonstrating their prowess, it’s the visualisation of a philosophy, it's artists learning from each other and competing with each other. Art movements and styles have philosophies behind them. Romanticism emphasised intense emotion, especially fear, terror, and awe along with the sublime. Vorticism extolled industry, speed, and geometric style. While the Baroque championed the complexity, contrast, and deep colour in opposition to Protestant art of the 17th century. Movements lasted a century, decades, or only a few years. There is purpose to them. But with AI image generators that is lost. There isn't anything behind what Midjourney, and its ilk, generates. It may look cool but it won't mean anything and that is the death of art.
Experience. Meaning. Depth. Awe. All lost to a machine incapable of feeling anything. Lost because 'it's cool bro.' As Wei Liu and Feng Tao write:
When AI art accelerates the artistic process, people’s expectations and experiences of art are reduced. This may not be due to a lack of depth in AI art itself, but rather to the fact that people are used to a ‘short experience/short memory’ mode of experience and are unwilling or unable to spend more energy on art content.
How does the “AI art” have depth when the prompter has given little and the AI even less? What depth can be created, what meaning can be instilled? I don't mean more time spent = more depth, though that can be the case. I mean what are the artists ideas, their viewpoint, the story being told through their work. The AI has none of that because it can't, it is a merely layers of algorithm. The prompter also cannot instil that in the image generated because of their lack of agency. Having an idea and typing it out is very different to bringing the idea to fruition. Nuance, learning, intention, all lost. Everything that matters is lost when the AI is involved.
Where the authors misstep with depth they get right with “short experience/short memory” and the question of depth vanishes. It doesn't matter if the AI or prompter produces depth if the viewer isn't bothering to investigate, reflect on, or simply sit and admire the work. A quick dopamine hit from your image on a Twitter doomscroll and on to the next one in #AIart. Hundreds of images never touching the viewer long enough to be remembered beyond breakfast. This inability to consider, to slow down, to appreciate and experience deeply is crushing humanity. The world accelerates and people become more and more depressed. The more experiences we cram in our eyeballs, our ears, and down our mouths the less we care, the less we remember, and the more miserable we become. AI “art” is merely another tempo increase in accelerationism, one we keep choosing to feed and one that will inevitably breakdown either in an explosion or a whimper.
Wei Liu and Feng Tao go on:
Kenneth Turan, the Los Angeles Times movie critic, is full of complaints about this: “the summer movies have been tailored more and more to the mindlessness often associated with the tastes of young males”; moreover, “(it) has as much interest in the human condition as a stone” [40] (pp. 5–8). Despite this, these popcorn movies continue to sell well and audiences, as a result, do not care. The reason for this is, in our opinion, that while young audiences may not have lost their taste, time constraints do not allow them to demand more quality content. When people become accustomed to the pay-as-you-go approach to life consumption, they follow the same stereotype in art consumption. In regard to the understanding of artworks, depth means time; therefore, when time is scarce, one loses patience with depth. Art in the age of acceleration is similar to a dessert: too much of it becomes boring but people are happy to try a little bit of everything. The rapid production capacity of AI art satisfies the demand for immediate and increasing varieties in terms of “taste”, such that people do not care about depth, but only about the richness of the experience, even if it is a short experience/short memory.
Film, music, TV, and games have suffered immensely from the slush-pipe of 'content' creation. And now the fine arts can join them. Audiences haven't lost their taste the vast majority never had a chance to develop one. You can serve up anything and people will clap like seals. Just look to the latest Disney+ or Netflix disaster show. There is more to be said on “time constraints” and “pay-as-you-go approach to life consumption” that is not in the purview of this essay.
“Depth means time.” Yes it does, certainly with regard to viewing art (or reading). You must invest time to get a worthwhile experience back. The statement “too much of it becomes boring” is false. It blames the art when it should be pointing to the acceleration. The ability to go slow in a fast paced world is invaluable. People are bored because they cannot think, they cannot consider, they cannot reflect, it is not that the people Wei and Feng are describing are thinking faster it is that they aren't thinking at all. They don't believe they have the time. Their faculties warped by a steady stream of 1 minute (if that) TikToks that fry the brain and corrupt the soul. Less and less people wish to wander around their local art gallery for an hour but will willingly stare into their hand-sized voids for an hour on the sofa. The time isn't the issue, the lack of depth is – both in capability of thought and desire to appreciate the sublime.
In fact, when AI artworks are accused of lacking depth, perhaps only one side of the coin is being seen—in the age of acceleration, depth may become a barrier to appreciation and circulation. AI art “becomes a glorified version of candy crush that seductively maims our bodies and brains into submission and acquiescence” [41] (p. 76). In short, AI art does speed up the art process, possibly to the detriment of the depth of the work; however, it fits the aesthetic needs of a time-poor reality.
What is meant by “barrier to appreciation and circulation”? It seems to me, being AI generated images are digital, this is a social media argument. One of internet points and impressions. Sure an AI generated image may get millions of views, hell billions of views, but the chance of it being remembered is almost nil. There is such a glut of “content” that aiming for acclaim shouldn't be the goal. Faster is not better and more likes is not better either. Focus on the work, on the skill, the meaning, the intent, and you will be far better served. We may live in a “time-poor reality” (of our own making) but I refuse to live that way. Faster is not better.
Thus, in the face of the information bomb, despite all the stimuli one receives, one does not have a long experience of time; rather, they receive only a short experience of time and the accompanying forgetfulness. As such, one is subsequently caught up in the panic of time being silently consumed.
Precisely. Short experiences create the illusion of running out of time. Slow down. Turn off your notifications, throw away the smart watch, mute your emails. Sit down and read for an hour. Go to an art gallery for an afternoon. Sit in the sun drinking a coffee and watch the world go by. Slow down, your life depends on it.
AI image generators fit naturally into our current hellworld but will only exacerbate the problems already present. Some artists will suffer, mostly in the commercial worlds of book covers, concept art, video game assets, and much of this is already happening. It will not stop there. With 'Nothing, Forever' demonstrating the ability for AI to produce a whole show on its own. While it is janky, bizarre, and dull it showed a glimpse of the future that could be, if we choose it. A world of AI generated mediocrity from screen to screen. Traditional artists, painting with brush and drawing with graphite, will weather the storm better as you cannot print the texture, colour, and feel of those works. Though Ai-Da is coming for us too.
With mechanisation comes the loss of aura, the essence, the soul, of the art work. Walter Benjamin writes:
The situations into which the product of mechanical reproduction can be brought may not touch the actual work of art, yet the quality of its presence is always depreciated.
One might subsume the eliminated element in the term “aura” and go on to say: that which withers in the age of mechanical reproduction is the aura of the work of art.
These two processes lead to a tremendous shattering of tradition which is the obverse of the contemporary crisis and renewal of mankind. Both processes are intimately connected with the contemporary mass movements... Its social significance, particularly in its most positive form, is inconceivable without its destructive, cathartic aspect, that is, the liquidation of the traditional value of the cultural heritage...
Writing in 1935 Benjamin cuts to the heart of the matter. Mechanisation obliterates difference in favour of mass society and mass society which requires a mass product to satiate its mass desire. No longer will small, localised, folk art be tolerated. Find “the best” and copy it ad nauseam. Share it on social media and let the like button take you to the top. Art, and artisanal work, once had meaning embedded in time and place. Localised cultures formed through the environment around them. No longer. With mass society all is made rootless and banal. An ocean of 'content' an inch deep. Mechanisation meant an artwork that existed alone (or perhaps with a few inexact and thus unique copies), that people had to pilgrimage to see, could now be printed and shared wherever. The aura was lost, destroyed in fact. Its significance seemingly maximised but in fact minimised.
The uniqueness of a work of art is inseparable from its being imbedded in the fabric of tradition. This tradition itself is thoroughly alive and extremely changeable. An ancient statue of Venus, for example, stood in a different traditional context with the Greeks, who made it an object of veneration, than with the clerics of the Middle Ages, who viewed it as an ominous idol. Both of them, however, were equally confronted with its uniqueness, that is, its aura.
Unique and particular cultures have different interpretations of the same artwork because the aura, essence, soul, is intact. It speaks but the words sound different to disparate listeners. AI, for all it's impressiveness, lacks any aura. It does not speak.
We know that the earliest art works originated in the service of a ritual—first the magical, then the religious kind.
Art never lost this ritual element, nor should it. The direction may have changed but art is still ritual, it is still a communion between material and spiritual, of the seen and unseen. AI “art” perverts this, flattens it down to mere material, and eradicates the nature of art. But this is easily undone, close the browser window and pick up a pencil. Draw what you see rather than what the programme predicts you are seeing.
In a few years this conversation may seem as naïve as asking 'is photography art?' did in Walter Benjamin's day.
Earlier much futile thought had been devoted to the question of whether photography is an art. The primary question—whether the very invention of photography had not transformed the entire nature of art—was not raised.
Will AI change the nature of art? Maybe. Did photography change the nature of art? Yes and not just in obvious ways like people rarely sitting for painted portraits but also in what art was produced. Painting, and other mediums, were influenced by photography such as Dynamism of a Dog on a Leash by Futurist Giacomo Balla.
Here the technology influenced the art and the work remains a product of the artist's mind. Whether it is art or not blurs with the photographs themselves. Photography is considered an art form by most, though dissent remains.
While photography lacks the sublime and transcendent, being a literal interpretation of the world rather than a human one, it still requires a person to spot a scene and capture it. The distinction between photography and AI is the role of the human is. With AI image generators the role has been diminished to a mere prompter (and occasional editor). Midjourney, et al, is the slow grinding down of the human into nothing. Plug ChatGPT into DALL-E and the human won't ever have to partake. Perhaps the results will be as banal and bizarre and barren as ‘Nothing, Forever’.
But is this all just techno-negativism? Neo-Luddism? Like the painters of old complaining about photography? Yes. Because they were right. When we utilise technology and technic we aim to expand our faculties. The paintbrush helps us paint, the rubber enables us to undo mistakes, the notebook allows us to remember that which we might forget. When we use a camera the intention is often to immortalise an important event that we are afraid of forgetting. Tangentially it can be used to capture impressive beauty. What does Midjourney and its ilk do then, does it expand the human imagination as David Holz claims?
[Rob Salkowitz for Forbes:] What is Midjourney’s mission?
[David Holz:]We like to say we’re trying to expand the imaginative powers of the human species. The goal is to make humans more imaginative, not make imaginative machines, which I think is an important distinction.
No. It usurps it. Midjourney is not a notebook extending your memory, it is a machine robbing you of your imagination. Let us take an example from Extended-Mind Thesis. Taking David Chalmers 3rd example (p.2) we can swap out rotating objects and insert AI image generation:
(3) Sometime in the cyberpunk future, a person sits in front of a similar computer screen. This agent, however, has the benefit of a neural implant which can perform the rotation operation as fast as the computer in the previous example. The agent must still choose which internal resource to use (the implant or the good old fashioned mental rotation), as each resource makes different demands on attention and other concurrent brain activity.
Becomes:
Sometime in the cyberpunk future, a person sits in front of a computer screen. This agent has the benefit of a neural implant which can read their thoughts. The computer will interpret the thought in text-form and produce an image based on it. The agent must still choose whether to “speak” to the computer or pick up a pencil with their hand and draw on paper.
The addition of the neural implant makes little difference as the process is effectively the same. Have a thought, translate to text, insert into AI image generator, image pops out. We can go further though:
Sometime in the cyberpunk future, a person sits in front of a computer screen. This agent has the benefit of a neural implant which can read their artistic thoughts. The computer will interpret the artistic thought and produce an image accurate to the thought.
Here we have gotten rid of the text-to-image part and jumped straight to mind-reading, which is not as unlikely as you'd think.
Would this mind-reading computer now be producing art belonging to the agent with the neural implant? No. The computer is once again doing the work – producing the image – and not the human, who is merely having an idea. The computer lacks the viewpoint of the person and as any artist would know the final outcome of a project is rarely identical to the original conception. Take this essay for example, over the 2 months of reading and writing it has changed a lot. I have included articles, papers, and paintings I didn't think to at the start (e.g. The Extended Mind and Dynamism of a Dog on a Leash) and much of what I made notes on didn't make the cut (I have 24'571 words of notes). Art, and all work humans do, evolves as it is created. The scope changes, the aim changes, hell the answer to a question changes. When I first imagined this essay I thought I would be arguing for AI “art” being art. The time spent reading, discussing, and thinking about the question allowed the idea to grow, develop, and mature. This time element is lacking in AI and creates shallow experiences, as discussed before. Not for the AI but for the human. It doesn't matter how fast the AI is, the human remains slow – “prosthetic” extending their mind or not.
There is an obvious difference between using a notebook to jot down an idea and using an AI to create your work for you. In this view, AI image generators are a prosthesis to our own creative abilities. Escalating the speed of art, the time spent enjoying and contemplating art, and the value of art itself. If a million images can be produced at the click of a button then the “art” lacks aura. The “art” it produces is dead. The technology doesn't help you, it doesn't improve your life, it negates it. Actively harming you.
In terms of aesthetic experience itself, the aesthetic structure of the masses has changed dramatically due to a lack of time, whereby there is a tendency to chase short, shallow experiences, a lack of experiential stimulation patterns, and a loss of patience with the pursuit of depth and value in artworks. The faster production capacity, quicker creation methods, and almost inexhaustible creative drive of AI art cater to the demand for aesthetic experience in a society accelerated by time scarcity, which may not be a wholly good phenomenon, however it is an indisputable fact. (Liu and Tao, 2022)
Accelerationism is happening but it is up to you whether or not you go along for the ride. Declining the wild ride of modernity is difficult, as difficult as riding it. There is an element of being left behind, especially if these technologies catch on in your industry, but for art it is important to remember one of its purposes, perhaps its main purpose. To see the world through the artist's eyes. Prompting a machine to create an image does not do that, that creates pictures that aren't art. They may be beautiful, they may be inspiring, etc., but these qualities are not unique to art.
It does not matter how close to the human mind the AI becomes, as The Distributist says:
...is a perfect simulation of plant photosynthesis equivalent to actual photosynthesis? Almost all intelligent people would say “no”. But then why do so many of them now think they can achieve immortality by uploading their brains onto the internet?
A simulation of a thing is not the thing itself. I assumed this was obvious so why do people think AI “art” is art? Why are people so eager to offer up their humanity on the altar of machines and technic? In the attempt to control nature through technology we will destroy ourselves, and not in some nuclear explosion, but in a whimper of following the orders and diktats given to us by those small devices in our pockets and the whims of our workstations. This is already happening. Watch people act when their email pings or an app pushes a notification to their phone's lock screen and their attention is distracted. The entirety of a work day in some Laptop Class job could be composed of 8 hours of distraction with never more than a few minutes spent engaging with a singular task. We have already become subordinate to our devices, our technology, and AI “art” is a logical step on this “ladder of progress”. The Distributist continues:
C.S. Lewis’s masterpiece The Abolition of Man. In Lewis's view, the conquest of nature will continue until the point where technology conquerors human nature itself, and the feedback-loop closes. Therefore, all subsequent generations will become the playthings (or slaves) of whatever intelligence exists in the unique historical moment when the creation of new sentient creatures becomes totally manipulable, an apocalyptic technological singularity controlled by an intelligence that possesses no connection to humanity or the function of ordinary morality.
This future is avoidable. You have a choice; remain human or become something less. A thrall to the machine. A machine which is not understood, we do not own or control, and which is rotting our ability to focus, think, and appreciate.
While the mass man will be suckered into AI “art” and other technological temptations, others will not.
Put starkly, and in a more “Carylylian” way, it takes an aristocratic temperament to rule the world. And men trained by robots, to think like robots will never be anything more than robots. Which, true to the etymology of “robot”, is just a longer way of saying men trained to be slaves will always remain slaves. For what else is a slave than a being that cannot take ownership over its own future? And isn’t that the true threat of modernity? That the human race will submit to enslavement by forces that aren’t even conscious?
I don’t think that this is our fate. The slave and the cog will naturally be subservient to the spirit of the aristocrat, even the aristocratic peasant.
The artist will continue to create art with their own wit and skill. Mastering the pencil and paintbrush through years of practice to create art that transcends the material world and offers a view into the sublime as well as the soul of the artist. This aristocratic artist will not care for the AI nor see it as a threat because they understand that they embark on a greater voyage than the simple production of pretty baubles. Art is dead, long live Art.