”Sold! Mario Klingemann’s pioneering Artificial Intelligence artwork sells to an online bidder for £40,000 in the artist’s auction debut. The race for AI generated art moving into the space of museums and galleries seems to get traction.”¹
More and more pieces of art emerge on the internet and show us the capabilities of artificial intelligence and their “creativity,” being it machines that write poems or rendering artworks. Are those artificial explorations, pieces of art or tech demonstrations? Should we classify pieces of art that materialized through or in collaboration with artificial intelligence as art? Does this AI-Art belong in a museum or gallery?
In 2018 the auction house Christie’s for the first time offered a work of art created by an algorithm. The AI-Artwork “Portrait of Edmond Belamy” sold for $432,000 and was created by Obvious, a Paris based collective. With the auction, the conversation shifted to the role of the author and creator. Obvious famously reused a GAN or generative adversarial network from Robbie Barrat. Who counts as the real creator or artist of “Portrait of Edmond Belamy,” Obvious, Robbie Barrat or the GAN itself?
As a result of the auction, Christie’s offers us a possible definition of AI-Art with their statement about the portrait: ”This portrait, however, is not the product of a human mind. It was created by an artificial intelligence, an algorithm defined by that algebraic formula with its many parentheses.”
With every technology that we use for creativity, the same questions about originality, authorship, and dependence on machines as a creator rise. Should we consider I-Art a new tool or medium?
On Artificial Intelligence
Today nothing seems more normal than viewing photography in a museum or a gallery. Photography had to establish itself from being more than a replacement of a painting, to become art, through time. When photography emerged the job of a painter had to face direct competition. As George Eastman founder of Kodak claimed in 1888 with the advertising slogan: “You Press the Button, We Do the Rest,” the work previously done by a painter was rendered obsolete with a simple click.
With the introduction of the camera, the portrait and landscape painter had to accept his handicap to capture reality instantly and righteous. The painting was a unique handmade original interpretation, the photograph an instant repeatable reflection of reality. The emergence of photography allowed the art of painting to become surreal, cubist and modern, while photography took over the job to capture and documenting reality.
“The machine gun approach to photography by which many negatives are made with the hope that one will be good is fatal to serious results. To take a good photograph runs the common claim, one must already see it. That is, the image must exist in the photographer’s mind at or before the moment when the negative is exposed.”
(Susan Sontag, On Photography)
The paintbrush is as much a tool for the painter as the camera is for the photographer. For now, AI-Art itself is a tool for the artist or author too. Decisions are made by a human being, from choosing an algorithm, selecting a training data set, favoring outputs to judging the representation of the outcomes. AI-Art, interim, requires a minimum amount of human creative input or decision making to allow creation to happen.
Photography disrupted the world of painting, as Artificial Intelligence will do with art today. Photography had powers that no other image-system has ever enjoyed, the same way Artificial Intelligence has skills that no other image-system has ever experienced. The same logic will be right for any following form of AI-Art. In its new forms, AI-Art will show us art as we have not seen it before.
Is this original?
How might we recognize AI-Art? Do we require new norms of aesthetics, fairness, and authorship to criticize and evaluate AI-Art? Currently, there seems to be no clear point of view yet.
Lately more and more galleries and museum picking up artworks that were created through or in collaboration with AI. And with that, they take a leading role in forming and clarifying the nature of AI-Art. Their selections will shape our understanding of how we will evaluate, look and perceive AI-Art.
The work of AI today is directly linked to a human actor in the same way photography is. The shock and novelty of AI-Art will wear off as with every technology that we added to our creative processes; it will dissolve into the background and become part of museums and galleries.
Trying to force a definition on AI-Art as a new form of art will always, similar to the discourse of photography and painting as art, be a push to hold some line. A line that will shift move and will be hard to hold. Any attempt to restrict AI-Art to particular subjects or techniques no matter how successful they are will be challenged and eventually collapse. For now, AI-Art is, as photography was, a new medium that enables to replace, revisit, push beyond and invent other media, shapes, forms, movements, and ideologies.
You Press The Button, AI Does The Rest.