Is artificial intelligence the future of art?

For many, they are the next big thing in art – digital images of jellyfish swishing and blurring in a deep pink sea, or dozens of butterflies merging into one. single creature.

Argentinian artist Sofia Crespo, who creates works with the help of artificial intelligence, is part of the “visual arts” movement, where humans create rules for computers, following that uses algorithms to generate new forms, ideas, and patterns.

The field has begun to attract great interest from art collectors – and even fetches even greater prices at auction.

American artist and programmer Robbie Barrat – a prodigy still only 22 – sold a work called “Nude Portrait #7Frame #64” at Sotheby’s in March for £630,000 ($821,000). ).

That comes nearly four years after French collective Obvious sold a piece at Christie’s titled “Edmond de Belamy” – largely based on Barrat’s code – for $432,500.

A ballet with machines

Collector Jason Bailey told AFP that synthetic art was “like a ballet between man and machine”.

But the early landscape may have been on the verge of a major shake-up, as tech companies start releasing AI tools that can generate lifelike photos in seconds.

Artists in Germany and the United States made a mark in computer generated art throughout the 1960s.

The V&A Museum in London holds a collection dating back more than half a century, one of which is a 1968 work by German artist Georg Nees called “Plastik 1”.

Nees used a random number generator to create a geometric design for his sculpture.

Computer ‘babysitting’

Today, digital artists work with supercomputers and systems known as Generic Adversarial Networks (GANs) to create images far more complex than anything Nees could have dreamed of. .

A GAN is a collection of competing AIs – one creates an image from the instructions it’s given, the other acts as a gatekeeper, judging whether the output is correct.

If it detects an error, it will resend the image for editing and the first AI will come back to life in the moment trying to beat the player.

But artists like Crespo and Barrat emphasize that the artist remains at the heart of the process, even if their working methods are not traditional.

“When I work this way, I’m not creating images. I’m creating a system that can generate images,” Barrat told AFP.

Crespo said she thinks her AI machine will be a true “collaborator”, but in reality, it is very difficult to get even a single line of code to produce the desired results.

She said it was more like “babysitting” with the machine.

Tech companies are now hoping to bring a piece of this rare action to the average consumer.

Google and Open AI are both touting the value of new tools that they claim offer realism and creativity without coding skills.

Enter ‘transformer’

They have replaced GANs with more user-friendly AI models called “transformers” that are capable of converting everyday speech into fluent images.

The Google Imagen website is filled with absurdist images created by tutorials like: “A little cactus in a straw hat and neon sunglasses in the Sahara Desert.”

Open AI boasts that its Dalle-2 engine can output any scenario in any artistic style from Flemish masters to Andy Warhol.

Although the emergence of AI has raised concerns about being replaced by machines in fields from customer care to journalism, artists see this development as an opportunity rather than a threat. .

Crespo tried out Dalle-2 and says it’s a “new level of visual generation in general” – although she prefers her GAN.

“I don’t usually need a very precise model to create my work, because I love when things look a bit hazy and unrecognizable,” she says.

Camille Lenglois of Paris’ Center Pompidou – Europe’s largest collection of contemporary art – also dismisses any suggestion that artists are about to be replaced by machines.

She told AFP that machines do not yet have “critical and creative capacities”, adding: “The ability to create realistic images does not make people artists.”