Rutkowski was surprised at first, but thought it could be a good way to reach new audiences. Then he tried searching for his name to see if a piece he was working on had been published. The online search returned work that had his name attached to it but was not his.
“It’s only a month. How about in a year? I probably won’t be able to get my job out there because [the internet] will be flooded with AI art,” says Rutkowski. “It’s worrying.”
Stability.AI, the company that built Stable Diffusion, trained the model on the LAION-5B dataset, compiled by the German nonprofit LAION. LAION compiled the data set and reduced it by filtering out watermarked images and those that were not aesthetic, such as images of logos, says Andy Baio, a technologist and author who downloaded and analyzed some of Stable Diffusion’s data. Baio analyzed 12 million of the 600 million images used to train the model and found that a large proportion came from third-party sites like Pinterest and art shopping sites like Fine Art America.
Much of Rutkowski’s artwork was scraped from ArtStation, a website where many artists upload their online portfolios. Its popularity as an AI application stems from a number of reasons.
First of all, his fantastical and ethereal style looks very cool. He is also prolific, and many of his illustrations are available online in high quality, so there are plenty of examples to choose from. An early text-to-image generator called Disco Diffusion offered Rutkowski as an example command.
Rutkowski has also added alt text in English when he uploads his work online. These descriptions of the images are useful for people with visual impairments who use screen reader software, and they help search engines rank the images as well. It also makes it easy to scrape, and the AI model knows which images are relevant to commands.