It is overwhelmed by AI-generated images, and some art communities ban it outright

A variety of robot images created by Stable Diffusion
Zoom / Variety of robot images generated by Stable Diffusion as found in Lexica search engine.

In the face of the huge amount of artwork generated by artificial intelligence, some online art communities have taken dramatic steps to ban or curb its presence on their sites, including NewgroundsAnd the ink splash artAnd the fur affinityAccording to Andy Bayo

Bayou, who was following Artificial Intelligence Art Ethics Closely on his blog, first notice and report the bans Friday. So far, the major art communities DeviantArt and ArtStation have not made any policy changes related to artificial intelligence, but some vocal artists On social media, they have complained about the amount of AI art they regularly view on those platforms as well.

Access to widely available jigsaw models such as Midjourney and stable spread It sparked an intense online battle between artists who view AI-assisted artwork as a form of plagiarism (more on that below) and passionate artists. embrace New creative tools.

Established artist communities stand at a difficult crossroads as they fear that non-AI artwork will be drowned out by the unlimited supply of AI-generated art, yet tools have also become remarkably popular among some of their members.

In banning art created by photomontage in his art portal, Newgrounds Wrote“We want to keep the focus on the art that people make and not overwhelm the art portal with computer generated art.”

Fur Affinity noted concerns about the ethics of how photomontage models learn from existing artwork, writing“Our goal is to support artists and their content. We do not believe it is in the best interest of our community to allow AI-generated content on the site.”

These are just the latest moves in a rapidly evolving debate about how art communities (and art professionals) can adapt software that can produce unlimited beautiful artwork at a rate that no human being can function without tools.

Part of a broader debate on the ethics of art

A collection of non-AI artwork used to train Stable Diffusion, compiled by
Zoom / A collection of non-AI artwork used to train Stable Diffusion, compiled by

The current wave of photomontage tools allow users to write a written description (called a “prompt”) and output a matching image, almost like magic. The results often require cherry-picking and dedication to get just right, but with a skillfully crafted router, the results can mimic the work of human artists in sometimes stunning detail.

The most successful routers often refer to current artists and art sites by name but rarely alone. Professional artists can create new innovative stylistic combinations. For example, this is the router that created the robotic woman in the center of the image at the top of this article at Stable Diffusion:

Beautiful cry! Female mechanical robot!And the half a pictureAnd the Complex detailed environmentAnd the photo-realistic!And the ComplexAnd the elegantAnd the very detailedAnd the digital paintingAnd the ArtstationAnd the concept artAnd the smoothAnd the intense focusAnd the ClarificationAnd the Art by Artgerm, Greg Rutkowski and Alphonse Mucha (Seed 79409656)

The most popular photo montages use the . extension latent diffusion technology To create new artwork by analyzing billions of images. In the case of Stable Diffusion, these images are sourced directly from the Internet, courtesy Lion-5 b Database. (Images online often come with descriptions, which are ideal to train artificial intelligence models.)

Recently, Bayou researcher and artificial intelligence Simon Willison collected data on more than 12 million images in LAION-5B and created search tool It allows users to peek into a small but representative segment of the much larger group. (You can also search LAION5B’s photo collection for artwork – or even your name –in show Hosted on Github.)

A search for Jeremy Lipking, a live-action fine art illustrator, returns results in the LAION dataset.
Zoom / A search for Jeremy Lipking, a live-action fine art illustrator, returns results in the LAION dataset.

Ars Technica

A few weeks ago, some artists started discovering their artwork in the Stable Diffusion dataset, and they are We weren’t happy about that. Charlie Warzel He wrote a detailed report About those reactions to The Atlantic last week. With battle lines firmly drawn in the sand and New creative tools for artificial intelligence It’s coming out steadily, and this debate is likely to continue for some time to come.

Leave a Comment