Boris Eldagsen and the AI art dilemma

The winner of a Sony World Photography Award has rejected the prize after it emerged that he created his image using artificial intelligence (AI).

German artist Boris Eldagsen won the creative category of the prestigious competition with a black and white computer-generated portrait of two women, titled Pseudomnesia: The Electrician.

During the ceremony, Eldagsen appeared uninvited on stage, telling the audience that AI images are not photographs and therefore his image should not have won. He later berated the competition’s judges for failing to distinguish between the two. “I applied as a cheeky monkey, to find out if the competitions are prepared for AI images to enter,” Eldagsen wrote on his website. “They are not”.

Subscribe to The Week

Escape your echo chamber. Get the facts behind the news, plus analysis from multiple perspectives.

SUBSCRIBE & SAVE

Sign up for The Week's Free Newsletters

From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.

From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.

Sign up

A spokesperson for the award accused Eldagsen of “deliberate attempts at misleading us”, said The Art Newspaper.

‘Machines have crossed a threshold’

“Our machines have crossed a threshold,” declared Kevin Kelly in Wired. “All our lives, we have been reassured that computers were incapable of being truly creative. Yet, suddenly, millions of people are now using a new breed of AIs to generate stunning, never-before-seen pictures.”

Eldagsen’s image is an “immensely evocative scene”, agreed Zoe Williams in The Guardian. The judges may feel like they were conned, but “maybe, when done well enough, AI images can’t be distinguished from photography by anybody”.

Eldagsen’s image was created using DALL-E 2, an image generator developed by OpenAI, a San Francisco-based company. Platforms like DALL-E 2 require prompts, which are a set of instructions from a user, to create their image.

Advocates say you still have to be creatively minded to produce these prompts. “Two thirds of the prompts are only good if you have knowledge and skills, when you know how photography works, when you know art history,” said Eldagsen.

Besides, artists have always used past work to create new work. AI systems like DALL-E are “just another tool”, Steven Sacks, a gallery owner, told CNN Business.

‘A sheer lack of effort’

Controversy over new art-making technologies is “nothing new”, said The New York Times. Many painters “recoiled at the invention of the camera, which they saw as a debasement of human artistry”, the paper added. But now some artists complain that their work is being “scraped” by AI art generators without their consent.

Tim Flach, a world-renowned animal photographer and the president of the Association of Photographers, “is among those who feel ripped off and says artificial intelligence can easily imitate the style of his images”, according to Sky News. At the moment “these images are being generated by scraping our images, taking them off our websites, but there’s no remuneration there”, he said.

Some fear AI will put artists and photographers out of work, while trade bodies are demanding greater regulation. The ramifications could be even more troubling. “I think this [is] redefinition of what creativity is – there’s never been such a huge gap between the sheer lack of effort or work or anything going in and the huge sophistication that then results coming out,” comic-book artist Dave McKean told the broadcaster.

The real challenge could be combating the rise in fake news and fake images more generally.

“The question isn’t, ‘Who made this art, a human or a machine?’ or “‘an machine-art be real?’ concluded Zoe Williams in The Guardian. “It’s more fundamental: how much truth is there in my reality?”