Technology has been making our lives easier for centuries. But art and creativity have always been at the heart of humans and with good reason: machines are incapable of feeling, choosing, inventing. Without personality, emotions, cultural context, political struggles... How could they create? And what for?
In the 2000s, machines first entered the sphere of creating art as a medium. And for several months now, digital art has been making the most of the possibilities offered by AI and is now exploring the metaverse and NFTs. But it goes even further. Today, the development of sufficiently advanced artificial intelligence and other technologies, such as blockchain, make it possible for machines to create works of art from scratch. For the time being it still functions on human request, but for how long?
So how does digital art interact with new technologies? Can an AI be creative? What is the future for art and the profession of artist? Let's dive into the most exciting and terrifying things about Web3.
Web3 Refresher: What Is It Exactly?
Web3, also known as the "decentralised Web," is a new generation of the Internet. Compared to the two previous ones, its key is data immersion and decentralisation. This version of the Internet was designed to repair the shortcomings of Web 1 and 2, which concentrated control of the Internet in the hands of a few companies.
3D will be at the heart of Web3's graphics as it will be focused on a virtual environment. The two most important pillars of Web3 are NFTs and metaverse; these two technologies will make a great deal possible, especially for brands that want to communicate with original and immersive content.
What’s Happening to Art in Web3?
Digital art and the metaverse were expected to meet. As early as 2004, the curator of the Whitney Museum in New York, Christiane Paul, distinguished between art that uses the digital as a mere tool for creation — for example, photography, printing or music — and art that uses the digital as a medium in its own right. This form of art suggests that the digital, in the broadest sense of the term, becomes the material for the creation, transformation and/or functioning of a work. So, as soon as new possibilities open up (to host, sell or even create digital art), it rushes in.
Hosting Art In The Metaverse
Since the idea of this new Internet is to further immerse Internet users, the role of the metaverse is central. This new technology invented by Mark Zuckerberg's company (renamed Meta for the occasion), is a virtual universe in which people can interact with the environment and other users. It can be virtual or augmented and immerses users in a stimulating experience. It's a bit like the "Matrix" or the parallel universe experienced by Di Caprio's character in Christopher Nolan's "Inception."
What is revolutionary and new in the art world is that established artists are taking on this new space while playing with the medium by constructing digital art pieces made for it. For instance, at this summer's edition of Art Basel, German artist Albert Oehlen presented his augmented reality avatar in the metaverse. This parallel space seems to be a new playground for digital art, a new branch of art established in the 2000s.
Buying Stuff — And Art — Through NFTs
Thanks to the NFTs, the mechanics of acquiring and owning digital works has reached a new level. What exactly are NFTs?
This acronym, which stands for "non-fungible token," refers to a good or asset that can be exchanged for another good or asset of equal value. So it's no surprise that some well-known NFTs are crypto-currencies (e.g. bitcoin), since they are easily exchangeable. But we are also talking about works of art (digital art, films, music...) or digital products (for the record, the first SMS and the first Wikipedia page in history were sold as NFTs).
When you buy an NFT, you become its sole owner and you also receive a certificate of authenticity. To prove the uniqueness of an NFT, it must be digitised and registered on the blockchain.
Once digitised on the blockchain, the history of a work (purchases, resales, date of creation, name of creator, etc.) is recorded forever. This is why auction houses specialising in art, such as Christie's or Sotheby's, have adopted the technology behind NFTs.
Jeff Koons, one of the most prominent American artists in recent years, has pioneered the interaction of his work with the metaverse and NFTs. Moon Phases, a 2022 creation, was composed both of physical sculptures that he sold as NFTs and placed on the moon (in the metaverse).
Creating New Kinds Of Artworks With IA
Now, the two previous examples were peripheral to human-made art. But what if art was actually created by a machine? This is exactly what is happening thanks to new software, which allows digital art to be created via artificial intelligence. This is the case with tools like DALL-E 2, Craiyon, Disco Diffusion and the young Google Imagen.
How does it work? You describe what you want to create. For example, Donald Trump as a bagel. Or a drowning gorilla plane, Claude Monnet style. Or Sleeping Beauty's castle in horror movie style. Are you still with me?
AI is unique in that it can create intersections and synergies between things that have absolutely nothing to do with each other; it can quite naturally make mixtures that a human would not imagine or picture. AI has no historical or cultural background and no personal limitations; it does what it is told to do (for now). As John Mauriello, YouTuber on the By Design channel, states, “this technology is like a distortion portal that reinterprets all the human experience and shows you a new world."
So, what does it look like? I’ve tested Midjourney for you.
Sunshine on the street at the parade, psychedelic
A drowning gorilla plane, Claude Monnet style
Paris Hilton in the style of Caravage
Bagel Donald Trump
Sleeping Beauty’s Castle in a Horror Movie
Big Bang in the style of Joan Miro
A crying antdog
An animal that does not exist (original and 2 upscaled versions)
A funny joke. Quite poetic this one!
Several schooners flying underwater
Vintage theme underwater
Iceberg dark logo
The Big Question: Can AI Learn to Be Creative?
In the future, will works created by artificial intelligence and works by human artists be sold at the same price? Will there be a need for a dedicated property rights and what status would these works have? These are the questions we will have to ask ourselves if artificial intelligence continues gaining importance.
It is clear that it is capable of learning. John Mauriello explains: “I’ve dedicated my whole professional life creating designs. This IA created created more designs that I could do in several weeks. Now, to be fair, a lot of these designs are not finished concepts. (...) A lot of them probably could not even be manufactured. But if I wanted to get anything even close to this level of quality a year ago, it would have taken probably 100 times more time. (...) The fact that’s is this good that early tells me it’s going to be a revolutionary concept."
The question remains: an AI can improve its algorithm, becoming faster, more precise, more relevant. But can it be creative? In the opinion of Arvind Krishna, Senior Vice President of Hybrid Cloud and Director of IBM Research, we are still a long way off. “Teaching AI what’s melodic or beautiful is a challenge of a different kind since it is more subjective, but likely can be achieved. You can give AI a bunch of training data that says, ‘I consider this beautiful. I don’t consider this beautiful.’ And even though the concept of beauty may differ among humans, I believe the computer will be able to find a good range. Now, if you ask it to create something beautiful from scratch, I think that’s certainly a more distant and challenging frontier.”
To sharpen her eye for art, but not to become an artist in her own right.
And that is to be expected. Indeed, if we ask ourselves what makes up the substance of art, we notice elements that could a priori be compatible with creation by an AI (a touch, a technique...) but also other elements that are just as important, if not more so, that are specific to the artist's personality and choices. Their experience, sensitivity, the angle chosen, what they want to convey at any given moment.
A work of art is also an evolving process: unlike the work of an AI which would have a beginning, an end and a mechanical pre-constructed process to get there, an artist iterates on a work of art, whose creation takes the form of meanders, with sometimes, according to the artist, a good dose of improvisation, errors, changes, forced compromises... And this is what makes the work unique and sometimes spectacular, touching, impressive, rough or shocking.
The Future of the Arts
Let's start by saying that AI will only be able to create digital art, music and film for the time being. The day we connect AIs with machines capable of automating the creation of sculptures, paintings or automating the movements of mannequins and sets to perform their own play, this will become a global issue.
For now, AI brings both opportunities and risks for digital art, music and film, which must be contained.
With its ability to improve so much faster than humans and to bridge seemingly unrelated objects, AI can be a tremendous source of creativity for artists. A kind of infinite reservoir of surreal ideas, to stimulate human creativity.
Marcus de Sautoy, author of The Creativity Code, believes that artificial intelligence is indeed more likely to emerge as a collaborator than a competitor. In the opinion of this Harvard mathematics professor, one part of the art market will remain unchanged, while another will use this new tool to open up its creative possibilities, just as the camera or camcorder was once seized upon. An opinion shared by Mario Kligemann, a visual artist: “As an artist, on the one hand, you want control over your work. But at the same time, you also want interesting accidents to happen. Using AI allows me to find the right balance between the two."
And while a UK study has shown that artists' jobs won't be affected by the existence of AI, the technology is likely to raise issues that will require legislation in the medium term.
An AI was able to create an original song in the same style and with her exact vocal pitch by listening to Ella Fitzgerald's discography. While specialists have doubts about the ability of this piece to become a new jazz standard, it does raise questions about the ability of an AI to industrialise the copying of original visual, musical and cinematographic creations, at low cost and on an industrial scale. Will the sale of these works be banned? Should their creation be restricted? Might we introduce a notion of plagiarism via AI? Only time will tell.