Artists hate drawing AI

Swedish artist Simon Stellenhag is known for his work that combines the natural environment with futuristic robots, industrial machines and alien creatures. Upon learning that an artificial intelligence is being used to mimic his style, he experiences some of the same anxieties as the image itself.

The AI ​​demonstration was conducted by Andres Guadamuth, an intellectual property expert at the University of Sussex in England who studies the legal issues surrounding artworks created by artificial intelligence. Gudamuz used a tool called Midjourney to create a spooky image of Stellenbosch and post it on Twitter.

Enter AI to draw pictures in 30 seconds

According to Guadamuz, these works are intended to highlight the legal and ethical issues associated with image algorithms. Midjourney is one of many AI programs that can create images on demand using machine learning algorithms and millions of tagged photos. After training you will be able to draw with great precision in the individual style of each artist.

The researcher chose Stellenhag because the artist may be critical of the work of artificial intelligence and opposed the experiment, but insisted he didn’t want to offend the artist. Guadamuz believes that a copyright infringement lawsuit will fail because the art form is not protected.

Stellenhag admits that credit is part of art culture, but he still hates artificial intelligence because it represents “something that the world of technology will force society to do in the future”.

AI doesn’t see copied works and believes tools like Midjourney can be used to explore new artistic ideas. But the Swedish artist fears new technologies will help enrich the world of tech. “AI is the newest and most dangerous of these. She takes all of the artist’s work without permission and uses it as the basic raw material to create a rich, profitable product. The bosses,” said Stellenhag.

Algorithms have been used to create images for decades, but in 2021 a new era of AI painting begins. In January, OpenAI introduced Dall-E, which can draw images from text descriptions. copy it

In April, OpenAI released Dall-E 2 with the ability to create images, illustrations, and drawings that resemble human creations. The application will be made public, the company said. The inventions are also intended for commercial use.

OpenAI applies word filters and tools to detect potentially offensive photos. Many companies are developing the same AI, but may develop different usage rules. Meanwhile, more and more artists are asking about the ability of AI to reproduce itself as users are given different tools.

RJ Palmer, specializing in drawing fantasy creatures and movies. Detective Pikachu, he tried Dall-E 2 out of curiosity and is concerned about the impact of AI on his work. Palmer was shocked to see people using the Stable Diffusion tool to change the style of their paintings. Adding the artist’s name to the description.

David O’Reilly, an artist who criticized Del-In, believes the idea of ​​using AI to learn from past work to create new paintings and make money is wrong. “They don’t own refurbished products. They charge for using Google images,” he said.

Jonathan Lowe, CEO of Danish photo company Jumpstory, doesn’t understand how AI-generated images can be used for commercial purposes. “I’m interested in this technology, but I’m very concerned and skeptical.”

OpenAI spokeswoman Hannah Wong confirmed that the company’s AI rendering service is being used by a large number of artists and is gathering feedback from the art community during development. “Copyright laws have historically adapted to new technologies, and the same will happen with AI content. We will continue to work with artists and legislators to protect the rights of content creators,” she said.

diving ann (Consequences Wired)