A South Korean startup is holding a competition to fill one of the world’s first galleries for machine learning-generated art in a bid to foster a nascent artificial intelligence creativity scene in the country.
The company, Pulse9, which makes AI-powered graphics tools, is soliciting art pieces that make use of machine learning tech in some way—whether to produce an image out of whole cloth or restyle or supplement an artist’s work—through the end of September.
The project is a notable addition to a burgeoning global community of technologists, new media artists and other creatives who are exploring the bounds of machine creativity through art, spurred by recent research advances that have made AI-generated content more realistic and elaborate than ever.
The medium had perhaps its biggest mainstream breakthrough in 2018, when Christie’s Auction House sold its first piece of AI-generated art for nearly half a million dollars—a classical style painting of a fictional character named “Edmond de Belamy.” That was also the moment that inspired the team at Pulse 9, which had just launched an AI tool to help draw and color a Korean style of digital comic called webtoons earlier that year.
“We asked ourselves, ‘Could we also sell paintings?’ and we started looking for art platform companies to work with,” Pulse 9 spokesperson Yeongeun Park said.
The company teamed with an art platform called Art Together on a series of crowdfunded AI pieces that proved to be more popular than they had expected—one hit its goal a full week ahead of schedule—and the team began considering parlaying it into a bigger project.
“With great attention from the public and the good funding results, we gained confidence in pioneering the Korean AI art market,” Park said. “So, we eventually decided to open our own AI art gallery.”
The company acknowledges that questions of authorship and originality still hang over the concept of AI art but stresses that the gallery is about collaboration between humans and technology rather than AI simply replacing artists. Even pieces generated entirely by machines require a host of human touches, whether it’s curating a collection of visuals for training or adjusting training regimens to achieve a desired results.
“The theme of this competition is ‘Can AI art enhance human artistic creativity?’” Park said. “We hope that this competition will also be an opportunity to discover creative, competent and new artists who would like to engage AI tools as a new artistic medium in their artwork.”
The goal is to establish AIA Gallery as a well-recognized institution in the art world and educate people on the potential for AI-powered creativity. The organizers hope the process will also inspire other efforts and create an AI creativity hub in the country.
“Groups or communities of AI artists have formed and are gradually growing, especially overseas,” Park said. “In the case of Korea, the AI Art market has not been well-recognized yet, but we’ve been continuing to play our role with our own initiative.”
The AIA Gallery recently partnered with one of the leading startups in the new space, Playform, which is led by Rutgers University Art and AI Lab director Ahmed Elgammal (after learning about the company from an Adweek article).
Progress in generative AI creativity isn’t confined to the art world, either. Agencies have started to experiment with various AI-generated graphics in campaigns, and brands have filed a slew of patent applications around the central technology powering the revolution—a neural net structure called a generative adversarial network.
Credit: Google News