OpenAI Invites Artists to Try Sora, a New Artificial Intelligence Video Creation Tool
On March 26th, IT House reported that OpenAI recently invited a group of artists to test their new artificial intelligence video creation tool called Sora. With Sora, artists have created a series of stunning experimental short films, showcasing the power and novelty of artificial intelligence to the fullest extent.
Different from the familiar OpenAI chatbot ChatGPT and image generation platform DALL-E, Sora is currently not available to the public. This Monday, OpenAI published a blog post titled “Sora: First Impressions,” revealing the results of visual artists, designers, creative directors, and filmmakers using Sora.
These short films range from 20 seconds to a minute and a half, showcasing exquisite visuals with predominantly surreal and abstract content. OpenAI’s resident artist Alex Reben created a 20-second short film, presenting creative concepts that could potentially become his sculpture works. Creative director Josephine Miller produced a video blending models with translucent colored glass, creating mesmerizing yet enigmatic visuals.
Not all works are equally abstract and complex, though. The short film “Air Head” by the production company Shy Kids might be the most entertaining. It tells the story of a man with a balloon-like yellow head, reminiscent of a sci-fi version of the classic movie “The Red Balloon.” Walter Woodman from Shy Kids mentioned, “Sora not only generates photorealistic images but also creates entirely surreal objects, which is what excites us.”
However, there are even stranger works. Digital artist Don Allen Stevenson III’s piece “Beyond Our Reality” resembles a distorted version of a National Geographic documentary, showcasing bizarre creatures like a giraffe-flamingo hybrid, flying pigs, and catfish-eel hybrids – as if a mad scientist had spliced different animals to create new species.
IT House noticed that OpenAI and the artists did not disclose detailed instructions for creating these short films, nor did they share the process from initial conceptualization to final production. Did the artists simply input a description of the scene, style, and realism level, press enter, and generate the work? Or did it involve multiple iterations to seamlessly blend the balloon-headed man with his shoulders or make the initially peculiar rabbit-armadillo creature endearing?
OpenAI’s invitation for artists to try Sora is not surprising. The fields of art, film, and animation, where artists thrive, may be impacted by Sora’s powerful capabilities. However, most artists view Sora as a tool to speed up the creation of commercial works. As Josephine Miller expressed in the blog, “Sora’s rapid and high-quality conceptual creation challenges my creative process and enhances my storytelling abilities. It allows me to express my imagination with fewer technical constraints.”