The original AI-generated spiral medieval village captivates social media

On Sunday, a Reddit user named “Ugleh” posted an AI-generated image of a spiral-shaped medieval village that quickly attracted attention on social media for its remarkable geometric qualities. Subsequent posts garnered even more praise, including a tweet with over 145,000 likes. Ugleh created the images using Stable Diffusion and a steering technique called ControlNet.

Reactions to the online artwork ranged from wonder and amazement to respect for the development of something new in AI generative art. “Never seen photos like this. Something new in the art world,” wrote one user coolest I’ve ever seen. You were so good,” AI artist Kali Yuga wrote on X.

Perhaps most notably, Y-Combinator co-founder and frequent social media technology commentator Paul Graham wrote, “This was the point where AI-generated art passed the Turing test for me.” Although Graham referred to the Turing Test (which purports to test whether the behavior of a machine is indistinguishable from that of a human being) as a metaphor rather than literally, he was clearly impressed.

Not everyone was impressed, of course, that some X users attempted to separate the compositional elements of the AI-generated spiral village. “It’s nice, but there are a lot of decisions a human wouldn’t make,” wrote a graphic designer named Trent. “Many shadows are incorrect and putting the chimneys right above the windows doesn’t make sense. Zooming in also reveals the noise patterns revealing the art of AI.”

In June we covered a technique that used the AI ​​Stable Diffusion image synthesis model and ControlNet to create QR codes that looked like rich works of art, including those inspired by anime. Ugleh took the same neural network optimized to create those QR codes (which themselves are geometric shapes) and inserted simple images of spirals and checkerboard patterns into it instead.

If guided by the message, “Scene of a medieval village with busy streets and a castle in the distance (masterpiece:1.4), (best quality), (detailed),” ControlNet rendered scenes in which the artistic elements of the images match the perceptual shapes of spirals and checkerboards. In one image, clouds arch overhead and people stand in a gentle curve to match the guidance of the spiral In another, squares of clouds, hedges, building faces and a wagon make up a checkerboard-shaped scene.

The magic of ControlNet

So, how does it work? We have already talked often about Stable Diffusion. It is a neural network model trained on millions of images retrieved from the Internet. But the key here is ControlNet, which first appeared in a research paper titled “Adding Conditional Control to Text-to-Image Diffusion Models” by Lvmin Zhang, Anyi Rao, and Maneesh Agrawala in February 2023, and has quickly become popular in Stable Diffusion. Community.

Typically, a stable diffusion image is created using a text prompt (called text2image) or an image prompt (img2img). ControlNet introduces additional guidance that can take the form of information extracted from a source image, including pose detection, depth mapping, normal mapping, edge detection, and more. Using ControlNet, AI art generators can more closely replicate the shape or pose of a subject in an image.

Using ControlNet and similar tips, it’s easy to replicate Ugleh’s work, and others have done so with fun effects, including checkerboard anime characters, an animation, a “goatse” medieval village (surprisingly safe for work), and a version of medieval village of “The girl with the Pearl Earring.”

Despite massive attention and numerous offers to turn artworks into NFTs, Ugleh has chosen to keep a low profile for now. On who pioneered a new ControlNet technique.”

If you want to experiment with ControlNet, this site has a good tutorial. Additionally, Ugleh posted a step-by-step workflow on Imgur, including the spiral and checkerboard template files.

While the artwork is notable, current U.S. copyright policy suggests that the images do not meet the standards to receive copyright protection, so they may be in the public domain. While AI-generated artwork is still a controversial topic for many for ethical and legal reasons, creative enthusiasts continue to push the boundaries of what is possible for an unskilled or untrained professional using these new tools . It is not yet clear whether and how the law will ever recognize the necessary spark of human inspiration that makes works like these possible.