Joy Cardaño said she was ordered almost every week to create art inspired by anime. Now, she said, that work has almost stopped, with many online users seeming to gravitate to an art of artificial intelligence instead.
From Ghibli studio-The illustrations inspired by the doll and the action figure “starter», An explosion of images generated by AI in recent weeks has sparked a new wave of concern among artists Like Cardaño, who argue that the use of the SAPE the importance of trained artists and removes their commission opportunities.
“The people who use it (AI generators) should respect the artists,” said Cardaño, who goes by Joyblivion on Instagram, in an email, calling the trend “so contrary to ethics”. “Even if artists vary how they do not want their art to be used, they refuse to listen. I think that the one who uses it or plans to use it should look for its impact on the artistic community.”
Many in the artistic community echo the feeling while they continue to monitor the latest IA advances, including the recent deployment of the OPENAI GPT-4O, which can generate text, images and audio. Chatgpt users can generate images Using the model for free. The rest of its capacities are reserved for paid users, with Members’ prices from $ 20 per month.
Cardaño, 30, who is based in the Philippines, said that she was a full -time artist since she graduated from the University. She mainly sells her work on Inprnt, an online store. His parts ordered usually start at $ 100.
After seeing the virality of Ghibli’s trend, she went to Instagram to highlight her work passed in the hope of emptying people to pay art, instead.
“Studio Ghibli Fan Art that I drew with my own hands without needing AI,” she wrote in an article on April 1, accompanied by a sample of her work.
Several other artists have adopted the same approach, publishing their own images while condemning the use of AI and circulating a video of the co-founder of the Ghibli studio, Hayao Miyazaki, calling the art generated by Ai “an insult to life itself”. The artists maintained the same momentum with the trend of the start -up pack in the figure of action, by publishing in the same way as their own human manufacturing version takes the trend.
Many artists say that they are particularly concerned about the ability of AI to imitate certain styles, which artists have perfected for years.
Hollie Mennert, 37, an illustrator based in Los Angeles, said that she had met A model of AI that was trying to reproduce her artistic style In 2022 – The year Chatgpt was published.
After having received messages having informed her that someone had made a model based on his artistic style, she decided to take a look at herself. She was not impressed by what she saw. She said that the images produced by the model were similar to hers, but that they lacked emotion and were not something that she “would really draw”.
“What would you do if you take the time that you type prompts in AI and just try to draw something for yourself?” Said Mennert. “I know that it does not always use everyone, but I am so much more interested in seeing what people offer for themselves, because it is always more personal.”
Some artists, such as the illustrator and caricaturist Sarah Andersen, are a legal action.
A few years ago, Andersen, 32, ended up in a position similar to the Ghibli studio after seeing people online created false versions of its comics using AI technology. This led him and two other artists to pursue three IA companies in January 2023, alleging a copyright violation, saying that companies used his work to train models. The trial is still active.
“Credit, consent and compensation is what we would like about these models,” said Andersen. “This is what we are trying to achieve with the proceedings.”
The proliferation of art manufactured in AI finally weighs on many artists.
“If you are an artist who mainly works on commissions, you probably lose a lot of income because it is cheaper, easier and faster to use a model to create an image than to pay an artist,” she said. “And of course, this is a financial problem, because if you want to use the style of this particular artist, you can also do it, and you can do it without their consent, so it becomes a big financial problem.”
OPENAI and other technological companies have taken some measures to try to prevent their AI models from directly copying particular artists.
Openai said He chooses to adopt a “conservative approach” with this version of the AI image generation. With GPT-4O, the model can generate images that resemble the aesthetics of certain artists when the names of the artists are included in the prompt, but society has added “a refusal that triggers when a user tries to generate an image in the style of a living artist”.
“Our image generation tools are designed to support human creativity, not replace it, help anyone to explore ideas and express new ways,” an Openai spokesman said in a statement.
NBC News tested the model and found that he would refuse to create images that have reproduced the styles of Mengert and Andersen. He has always created images in the style of the Ghibli studio, but not when he was asked to create an image in the style of Miyazaki.
The CEO of Openai, Sam Altman, defended the trends of the art of AI in an interview with the content of content Varun Mayya Posted on YouTube this month.
“I think that the democratization of content creation was a big net victory for society,” he said when he was asked about his reflections on the reaction of artists. “It is not a complete victory; there are negative things on this subject, and it certainly did something about the art form. But I think that on the whole, it was a victory.”