In the majority of fiction that concerns artificial intelligence (AI), it replaces us in almost every industry. Often, only the artistic fields are left untouched. Usually, a robot or AI showing capabilities to create a work is seen as a completion, at the pinnacle of intelligence, considered almost human. Yet, AI and the arts are already flirting in reality, whether in music creation or visual arts, it is becoming more and more present.
However, on August 26, 2022, there was a rude shock when, at a visual arts competition in Colorado, the winning image was entirely designed by an AI. To arrive at his “Space Opera Theater,” Jason Allen only had to write a few words in a generator called Midjourney.
Writing to compose an image
The American businessman wanted to compete and will have succeeded in getting this image taking up some baroque codes. This result has aroused controversy as much among the artists present in this competition as art critics. Suddenly, Jason Allen was opening Pandora’s box by showing that, from now on, artificial intelligences are able to create unique illustrations in a few seconds that could very well be included in exhibitions.
For visual artists, this reality has, possibly, just sounded the death knell of their careers. How do we combat this “prompt art”? Indeed, there are now various software programs on the Internet such as Midjourney but also Dall-E or mage.space that create images in a matter of seconds simply by typing words.
This can be done by specifying a genre of painting or even an artist. The intelligence works on an artistic algorithm. It will fetch representations from its “memory” of objects, animals, people, styles, etc. It will then design an illustration from what it has unearthed and according to an approach decreed by the user or by his “creativity”. This Internet user, a draftsman in his spare time, shows how the intelligence manages to make several proposals in a few seconds and some turn out better than his sketches.
So, it’s entirely possible to get some stunning and convincing results in just a few words. If, for example, the application is well entrenched in popular culture, this allows for even more possibilities. Here’s one of the creations I made by typing “gandalf monet” on mage.space:
You’d almost think the French painter would have, in between paintings, made a portrait of the famous wizard from Tolkien’s Lord of the Rings. Yet, no, this was one of the most successful realizations that the AI offered me. Because yes, it took me several proposals before getting a satisfactory result. Several of the creations were not up to expectations. However, unlike a human who would have to start from the beginning and complete it in several hours or days, the algorithm only needs a few seconds to create something else.
Legitimate debates
This new approach raises a lot of questions. First, in terms of authorship: is it the AI or the one who composed the command to create the image? The legislation is quite clear on this point. As long as a human is responsible for the creation, he is the rights holder. The U.S. Copyright Office settled the issue in early 2022 when an AI creator tried to copyright the design conceived only by his artificial intelligence. Thus, not possible for an interface or algorithm to own any property.
Speaking of law, “prompt art” raises questions since artificial intelligences draw from other artists’ styles and works without asking their opinion. In fact, Quebec artists have been very surprised to see AIs already appropriating their way of painting. In fact, painters and sculptors are now checking sites to make sure algorithms haven’t started training with their artistic approach. Many of them are angry at these unintentional “borrowings” for which there is almost no legal recourse. Case law shows that inspirations cannot be considered plagiarism. This videographer explains it very well.
Besides, in the wake of the AI winning a contest, sites dedicated to visual art and image banks reject works composed by AIs. A bit of a harsh judgement: are artists using AI really not? Because looking at the single issue of “prompt art”, it is much harder than people may think to get the desired illustration.
The whole question of what is written, the order in which it is done, the precision of the order and the deep learning of artificial intelligence will lead to more or less successful achievements. So the skill is to find THE right combination. By the way, Jason Allen always refused to say what he wrote in Midjourney to get his award-winning image.
Even more so as some artists use the algorithmic tool in their work by machines or simply by quickly arranging photos to form another. So does that mean they are less of an artist by using a newer technology? It often seems that people consider intention to be artistic. The sensitivity of the one holding the pencil, brush, or pen remains at the very heart of creation as a whole. Is one less sensitive when typing on a keyboard in a will to create? The question deserves to be explored further.
Let’s not forget that the art world has always been conservative and looked down on new approaches. The piano was viewed with disdain at a time when everyone thought the harpsichord was the musical pinnacle. Musicians, many centuries later, were afraid that synthesizers would take away their jobs, since they could reproduce the sound of instruments. Yet this loss of jobs in music did not happen.
Certainly, the issue of “prompt art” raises many questions. Some believe that it would eliminate the need for art classes. Yet, we might counter that it is just a new method that deserves to be studied in an art curriculum. The tool comes with its moments of grace and its flaws like any other.
Tackling the issue with students will further analyze the possibilities, risks, and limitations of this technology that will be used in both good and bad ways. Any tool is neutral; it is the uses that give it meaning.
Photo credit: en.depositphotos.com
muzico LTD
Ursprungstrasse 1d
8044 Zürich
contact@muzico.ai
+41 78 905 12 65