Comprehensive coverage

Artificial intelligence, painting - and the bright future of art

Thanks to artificial intelligence engines like DallE and Midjourney, any snotty-nosed and future-talented child can produce impressive paintings in any style, shape and color. The debate revolves around the question of whether it is art

Duchamp's "Fountain". From Wikimedia
Duchamp's "Fountain". From Wikimedia

Jason Allen doesn't seem like a particularly annoying type. He has a thoughtful look in his eyes and an unkempt beard that instantly labels him as an artist-at-heart. This is a particularly ironic diagnosis, because in the last week Allen managed to stir up the international community of artists, especially the painters. The reason? He won first place in the Colorado State Drawing Contest, and did so with a drawing that artificial intelligence produced for him.

The reaction in the Twittersphere was not long in coming, and mainly obeyed the catastrophic lines expected of it.

"We are witnessing the death of art," one tweeter summed up the feelings of many on the subject, "dangling right in front of our eyes." [1]

One can understand the feelings of the human painters. I can testify from my personal experience that thanks to artificial intelligence engines like DallE and Midjourney, any snotty-nosed and future-untalented child can produce impressive paintings in any style, shape and color. All that is needed is to explain to the artificial intelligence what you want it to draw for you - and five seconds later, the drawing is already in your possession. You don't even need to know programming. You just have to explain yourself to the AI ​​well enough, and it will do the work for you.

Are we, therefore, really predicting the death of art?

To answer the question we first need to understand what art is.

In 1917, the French artist Marcel Duchamp presented his most famous work: "The Fountain". It is, simply, a urinal that the craftsman bought in a hardware store in New York and placed it upside down on a post. She so shocked the managers of the exhibition where she was supposed to be shown, that they chose to hide behind a curtain. It wasn't the vulgarity that bothered them, but the uncomfortable questions it raised: Can an artist present any object as art - even if it is nothing more than a simple variable? And if so, do every hammer, hoe and stove constitute works of art?

Duchamp's answer was a resounding yes. One of the critics of the work praised it with the following words -

"It doesn't matter whether the craftsman created the fountain with his own hands or not. he Select her. He took a simple object from life, placed it so that its significant utility disappeared under the new title and point of view - and created a new thought for that object."[2]

More than a hundred years have passed since the appearance of the "fountain", and it seems that the message of the offense was received and well understood. We can find in the most celebrated museums, works defined as "modern art" which are nothing more than simple objects arranged in extraordinary forms. On the walls of the homes of billions, you can see works consisting of paper clippings or paint spots sprayed as if randomly on the canvas. These are also accepted today as works of art.

What is art, then?

As you can understand from works like the "fountain", or the paint stains and the paper cuts, art does not require extraordinary technical skill. Many modern works of art today seem simple at first glance, but they evoke emotions and thought among the viewers - and thus their greatness.

Wikipedia sums up the term art well, in one definition that manages to include the cave paintings of the ancient man, the Mona Lisa and the "fountain" in one sentence. "Art is a wide variety of human activity, which amounts to a product that involves creative talent or rich imagination that indicates technical skill, beauty, emotional power or conceptual ideas."[3]

Art, in short, is simply anything that a person creates - and provided that behind the work there is a conscious desire to evoke feelings and emotions in the viewer.

And if this is the case, then we are currently entering an era of artistic abundance that has never been seen before.

Put yourself in the shoes of an ancient writer from a thousand years ago. Before he sits down to write, he has to purchase ink - an expensive commodity at the time. He has to fill the inkwell, dip the quill into it, and then think for long minutes before he dares to bring it closer to the page. He knows that from the moment he writes the words on the page, they will remain there forever. He will not be able to delete them - there is no undo button in the real world. After writing he will also need to let the ink dry, and maybe even sprinkle some fine sand to absorb the residue. Writing a single page can take a whole night of work.

And today? We open the computer, and almost without thinking produce a record like the one you are reading now.

The ease of writing is what brought us to today's world, where the number of published books increases every year. Not all of them are masterpieces, of course, but all of them provide a way for ordinary people to express their thoughts and ideas, to weave heroic and sinister plots and literally according to the definition - "to arouse in the viewer [or reader] feelings and emotions".

We are about to see a similar phenomenon taking place in the world of painting in the coming years. Each and every one will be able to produce pictures and drawings with the ease of thought - and they will do so. The world will be filled with art, and most of it will be bad. The number of lay artisans - those without real training - will be orders of magnitude greater than the number of artisans who were allowed to pass through the gates of the Academy of Arts. But among all the garbage, diamonds will also appear. People who would never enter the field of painting, will produce innovative and thought-provoking works of art, using techniques that will seem foreign and strange to us - just as Cubism and Surrealism would have seemed to our ancestors two hundred years ago.

And history proves that we will love them.

When the camera first appeared in the 19th century, it immediately became the loathing of painters. Many of them were making a living at the time by painting portraits and realized that they were going to lose this source of income in favor of the machine. A similar logic was also valid for painting landscape mirrors. Who needs the human painters anyway, if the camera is able to produce a more accurate reproduction of reality, in a much shorter time and at zero cost?

The simplest answer is that you really don't need them to reproduce reality Accurately. The painters who specialized in replicating portraits or landscapes on the canvas found themselves without work. Worse, they discovered that their work was no longer valued as it had been up to that time. The ancient Greeks praised the artists who could imitate nature so well that the birds would try to eat the fruits on the canvas. If those ancient artists had appeared in the middle of the twentieth century, they would have been ridiculed and scorned: why try to do something that any child can do, using a simple camera?

But humans continued to search for uniqueness: they wanted extraordinary visual art, the kind that even the camera could not provide. Some artists decided not to use the camera, and developed the strange styles we know and appreciate today: Cubism, Surrealism, Suprematism and Futurism (yes, there was such a movement), among others. Other artists used the camera in unusual ways that few could replicate: they spent years developing unique lenses, or inventing slow exposure techniques that required them to photograph the same scene for long nights. Or they would go on wild trips, and lie in wait for months just to get that one-perfect picture of a snow leopard looking deep into the eyes of a blushing, shy silkworm.

The art, in short, survived. It is more than an office: it is to the heart, it flourished and prospered. She enjoyed the works of countless people who could use the camera to capture experiences visually. Each such recorded experience was a small work of art in itself, with a unique message: a father walking with his children on a quiet street, a family hugging against the background of a birthday cake, a child winning the first football game of his life. But more than that, the professional artists, in their attempts to differentiate themselves from the masses, added and created new types of visual art that we can appreciate today as real works of art.

This history is about to repeat itself even in an era where artificial intelligence engines are able to produce images and drawings on demand. You can relax, artists: you will not have less work in the future.

In fact - and again, I am relying on history here - you will have even more work. And this is thanks to an extraordinary fact that we have learned in the last two hundred years: technology creates more new jobs than it eliminates.

The vast majority of us do not remember the era when bowling was a game without machines. It was a dark time in history, where after each roll a human worker was needed to rearrange the pins. Games took longer and required the presence of a human worker on each track. Bowling was only reserved for people who were able to pay for this extra work, and since there weren't many of them - there were only a (relatively) small number of bowling alleys.

Put yourself for a moment in the shoes of an economist from the beginning of the XNUMXth century, who came across an innovative invention: a bowling machine, which can pick up the pins that have fallen in each turn and put them back in their place automatically. Obviously, the machine can replace many workers in any bowling alley. Would you think it would decrease the number of jobs open to humans, or increase them?

The intuitive answer is of course that such machines will take the jobs of almost all the "bowling boys", and therefore will cause unemployment. This is also the correct answer, but only in the short term. The complete and surprising answer is that the bowling machines actually increased the number of jobs open to humans.

What happened? Simple: the machines dramatically lowered the operating costs of the bowling alleys. The managers of the venues preferred to use machines instead of people, thus saving on expenses - something that translated over time into a significant discount on the costs for the players as well. More people could play bowling, and it was easier than ever to open and maintain bowling alleys - and the result was that new alleys were opened all over the country. The number of theaters between 1955 and 1963 doubled, as did the number of players, which increased from three to approximately seven million.[4].

Would you say that even in all those new halls there were still no ball boys? are right But in every such hall there was a need for many other job holders: cleaners, managers, receptionists, accountants, repairmen, and much more. Eventually, the increasing mechanization of bowling meant that many new jobs were created. Much more than the number of jobs that were closed.

Bowling hall from the end of the 19th century. Link

I believe that we will see a similar phenomenon also thanks to painting through artificial intelligence. The fact that illustrations can be produced easily will leapfrog certain areas of the industry. For example, the board games, which have recently been renewed. Every board game today has dozens - even hundreds - of cards with unique illustrations. The result is that in order to release a new board game, you have to invest thousands of dollars in paying artists to produce all these drawings themselves. The same goes for many card games and computer games - except that there the costs can easily reach hundreds of thousands of dollars.

These three markets are about to undergo a dramatic disruption thanks to artificial intelligence. The cost of creating new games will decrease significantly, since it will be possible to produce hundreds of drawings at a very low cost. The result will be that more games will be released - and the production of each of them will provide jobs for many. Also for project managers and storytellers, but also for graphic designers and people who specialize in working with artificial intelligence engines and understand how to guide them to get the best results on the first try.

In the long run, AI engines will mean that there will be a greater need for artisans than ever before.

But - and this is important - the artisans will have to be ready for the new market, and will have to know how to use the new technologies.

Justin Reckling belongs to a new generation of painters - one that has appeared in recent months. Instead of using a brush - physical or virtual - Reckling paints with words. He sells his creations - verbal descriptions that explain to the artificial intelligence what it should draw - on the PromptBase website, which was opened specifically for this type of sales[5]. And he also makes a living from them at the same level as an average craftsman: barely. Every day he sells between three and five such descriptions, for a few dollars each. Luckily for him, it's just a hobby.

The descriptions Shrekling sells, also known as "prompts", are a new kind of art. The AI ​​engines rely on them to produce the final illustrations. Accurate prompts will yield the user images and illustrations that match exactly what he wants. Rougher prompts will result in less accurate results, and users may find themselves running several dozen attempts until reaching the perfect prompt that reflects their vision. It's understandable why people are willing to pay for Rackling's special prompts.

Creating these prompts is a work in progress. A relatively simple prompt can look like this, for example –

"White girl, 22 years old, Alisha Cuthbert mixed with Zdenya, techie girl, with braided hair of different colors, plasma rifle, sci-fi character concept, flawless, symmetrical, 8k, render octane, very detailed."

Caucasian girl, 22 years old, Elisha Cuthburt mixed with Zendaya, mechanic girl, with multicolored braided hair, Plasma rifle, SciFi character concept, no blemishes, symmetrical, 8k, octane render, highly detailed

Link to the picture on the PROMPT HERO website
Link to the image on the website On the PromptHero website

This, as mentioned, is a relatively short prompt. The longer prompts easily reach tens of words in length, with descriptions that look and read like art in themselves. And for good reason: they really are art. This is the first step in the new art: giving the right instructions to the artificial intelligence, so that it realizes the artist's vision.

This is just one example of the way in which the nature of the painters' work will change: they will spend less time illustrating and painting themselves, and more time managing the artificial intelligence. They will have to learn how to talk to her, how to explain to her what they want, how to choose the most successful products from the choices she shows them - and how to edit them with her help. Whoever knows how to do these things well will be hired and produce works of art left and right.

Fortunately for the painters, they arrive armed with the tools they need for this. The creative artificial intelligence engines have been trained on countless drawings, and can produce new versions - but you need to talk to them in the right language for that. The most talented prompt creators understand the difference between "rendering octane" and "watercolor colors", know how to instruct the engine in which drawing and coloring style to use, and know many works of art whose style the engine can learn - and successfully imitate.

In other words, the best painters become those who also know the theory and background behind the art. These give them the ability to get what they want from the engine quickly and efficiently.

The second advantage of the painters who would like to use artificial intelligence is that many of them know in advance what they want to produce. They have the background and experience necessary to visualize the finished image - and then they only have to describe it. Other mortals, with less experience in painting, will have more difficulty thinking ahead about the finished artistic vision, and will have to rely on the artificial intelligence to suggest ideas and direct them.

In short, if you studied art and painting, you can relax: you didn't waste your time. You too can have a job in the future, and the training you have acquired will serve you well in it. And on one condition: that you learn to work with artificial intelligence.

The thought that oppresses artisans lately is that creative artificial intelligence will take over their jobs. The truth is more complex: artificial intelligence is able to take on part of the painting work - the part that was considered the most difficult in our opinion so far - but artists also have a place in this revolution. They will guide and guide the artificial intelligence, help it produce the most successful paintings and images, and will only find themselves even more needed when the various markets that rely on such works of art leap forward. In the coming years we will enter a renaissance of comics, computer games and board and card games. Each of them will be based on pre-computer art - and the artists who know how to talk to artificial intelligence will be more important than ever.

My second happy prediction is that art will only become more diverse, rich and yes - also weirder. Everyone will be able to produce art, and new ideas for works will come from every direction, from every age and from every person. Just as the camera did not bring the art of painting to an end, but made painters look for new directions to differentiate themselves and their art, the same will happen to art following the popularity of creative artificial intelligence.

The best days of the artists are still ahead of them.

Now we just have to hope they understand it.


[1] https://twitter.com/GenelJumalon/status/1564651635602853889

[2] http://sdrc.lib.uiowa.edu/dada/blindman/2/05.htm

[3] https://en.wikipedia.org/wiki/Art

[4] https://www.theatlantic.com/technology/archive/2014/03/let-the-good-times-roll-the-incredible-bowling-bubble-of-the-1960s/359787/

[5] https://www.theverge.com/2022/9/2/23326868/dalle-midjourney-ai-promptbase-prompt-market-sales-artist-interview

More of the topic in Hayadan:

Skip to content