First photographers were creating portraits of people that don’t exist, now Aurel Manea has created a series of “landscape photos” using a new artificially intelligent (AI) software program called Stable Diffusion.
Manea tells PetaPixel that he has been blown away by what the London and Los Altos-based startup Stability AI has created.
“I can’t, as a landscape photographer myself, emphasize enough what these new technologies will mean for photography,” explains Manea.
“Of course, they are not real photos and they only resemble real places (for now, as the input data becomes larger and larger) but for most of the people that consume the images, it is only about the beauty of those images.”
Manea says that he has used DALLE-2 and it does “great images of people’s faces that look like photos.” But, he says, DALL-E fails to create landscape photography.
“It is probably because of the images the model was trained on and the politics of the creators to exclude contemporary landscape photographers,” he says.
Stable Diffusion, on the other hand, appears not to have such a problem which Manea puts down to having a wider image base than DALL-E.
“Just telling the AI something like ‘landscape photography by Marc Adamus, Glacial lake, sunset, dramatic lighting, mountains, clouds, beautiful’ gives instant pleasant looking photography-like images.”
“It is incredible that technology has got to this point where mere words produce such wonderful images (please check the Facebook group for more).”
Open Source AI Image Generator
Stable Diffusion is currently still in beta version, but differs from other AI image generators, such as DALL-E, because it has no content filter and is open source.
“Stable Diffusion will allow both researchers and soon the public to run this under a range of conditions, democratizing image generation,” Stability AI CEO and founder Emad Mostaque writes in a blog post.
“We look forward to the open ecosystem that will emerge around this and further models to truly explore the boundaries of latent space.”
Eyebrows have been raised about Stable Diffusion’s lack of safeguards, with Tech Crunch reporting that it “poses tricky ethical questions for the AI community”
AI image generators open up a world of possibilities which is what makes the technology so exciting. But, it could also open doors for crooks, frauds, and shysters.
Mostaque’s company, Stability AI, created LAION 5B, an open-source, 250-terabyte dataset containing 5.6 billion images scraped from the internet.
It also added a subset of two billion AI-filtered images that were ranked as “beautiful” by testers of Stable Diffusion.
Unlike DALL-E 2, which says that it owns the generations created. Stability AI does not assert rights over images created with Stable Diffusion and the images belong to the users who prompted them.
Image credits: All images by Aurel Manea.
Author: Matt Growcoot
Source: Petapixel
