Laura Rautjoki creates art using artificial intelligence, a rarity in the art world
Writer: Vesa-Matti Ruuska, published in YLE on November 16 at 5:32 PM.
The article was originally published in Finnish and translated into English by Kate Vass Galerie.
Source: yle.fi
While AI-generated images have taken social media by storm, their use in fine art remains limited, partly due to ethical concerns surrounding the technology.
The images featured in this article are from Laura Rautjoki's series The Image of a Woman, which explores the portrayal of Finnish women throughout photographic history. However, these are not traditional photographs—they were created using artificial intelligence. “I’ve intentionally left some of the AI’s mistakes visible. For example, some figures have disproportionate features,” Rautjoki explains at the Amiraalistonkatu Warehouses in Turku, where her works are on display until the end of November.
While many online AI tools enable anyone to generate images, Rautjoki’s creations have been exhibited in a museum in Finland and a gallery in Budapest. What distinguishes these works as art? According to Rautjoki, it’s her perspective as a trained artist and photographer with over a decade of experience. She uses AI as a tool to express her unique visual language. “I feel like I’ve truly found my voice through AI-generated imagery,” she says.
“Airy" is one of the early images from Rautjoki's AI project. "The image demonstrated that AI can bring even the wildest photographic visions to life," she says. Image: Laura Rautjoki
Researcher: The Final Result Must Be Controlled by Humans
Determining the artistic value of AI-generated works is not straightforward, says doctoral researcher Iikka Hauhio. One way to assess their value is to consider whether the final result is controlled by AI or by a human. For human control, three key conditions must be met: the ability to guide the AI, ensure the AI’s predictability, and allow the human to express themselves through it.
“A camera takes predictable photos, a pen draws predictable lines, and an instrument produces predictable sounds. AI, however, is often unpredictable,” explains Hauhio, who studies computational creativity and art-generating programs at the University of Helsinki.
The artistic value of AI-generated images is further challenged by the ease with which they can be mass-produced. Even if an individual image is aesthetically strong, its value diminishes when similar images can be created in unlimited quantities.
According to Hauhio, AI can indeed create art, but only when the final result remains clearly under human control. He uses music as an example: a musician might ask AI to generate samples—short audio clips—and then integrate these into a song. “In that case, the artistic value comes from the human’s mixing and composition, which gives the work its significance,” Hauhio explains.
Another approach is to use AI iteratively, generating multiple outputs from the same concept to refine the result. The human retains control by adjusting the inputs provided to the AI, shaping the final outcome. This method is also employed by Laura Rautjoki in her series. Below are examples of AI-generated images that, after input adjustments, became part of the work Getting Ready for School.
"They Know Me" depicts "prosperous women in the Forssa region in the early 1900s." These images are part of Rautjoki's thesis for her Master’s degree at Turku University of Applied Sciences. Image: Laura Rautjoki
Use of AI in Visual Arts Still Limited
The use of artificial intelligence in contemporary art remains limited, with only a small number of artists identifying as "AI artists," says Tuomo Rainio, a lecturer at the Academy of Fine Arts at the University of the Arts Helsinki. “For an artist, automating their work process isn’t particularly meaningful, yet much of technological development seems focused on this,” Rainio explains.
However, the creative fields are paying close attention to advancements in AI technology, particularly regarding its potential to complement artistic processes and materials. A major point of debate centers on copyright issues. AI relies on datasets collected from the internet to generate images, many of which include visual art and copyrighted materials. “AI is a contradictory technology. It recycles the lifework of individual artists,” Rainio observes.
Photographic artist Laura Rautjoki faced similar ethical challenges in her work. She initially planned to use copyright-free images from the Finna service, maintained by the National Library of Finland. However, this would have required uploading those images to the AI system. “That created a strong ethical dilemma. Even though the images are copyright-free, I didn’t want to feed them to the AI because I don’t know where the material ultimately ends up,” Rautjoki explains.
Instead, she chose to use an AI application, despite the inability to verify the origins of the generated images. Acknowledging these complexities, Rautjoki refrains from claiming full ownership of the results. “I’ve created the images in collaboration with AI,” she says.
“Getting Ready for School” – Laura Rautjoki aimed to create documentary-style images for the exhibition, inspired by her own experience of living in a student dormitory. According to Rautjoki, this image came together relatively quickly, although some works require dozens of iterations. Image: Laura Rautjoki