Microsoft-backed OpenAI is developing software that can generate minute-long videos based on text cues, the company announced on Thursday.
The program, known as Sora, is currently available for red teaming, which helps discover problems in the AI system and access by visual artists, designers, and filmmakers to provide feedback on the model, their website stated.
“Sora can generate complex scenes with multiple characters, specific types of motion, and accurate details of the subject and background,” the company said.
The existing model has flaws. It may struggle to simulate a complicated scene effectively and may not comprehend precise examples of cause and effect. For example, the company noted that a person may bite a cookie, yet the cookie may not show any biting marks afterward.
“We’ve also developed robust image classifiers that are used to review the frames of every video generated to help ensure that it adheres to our usage policies before it’s shown to the user,” it added.
Ariel Ben Solomon is the Growth and Strategy manager at Ecomhunt. He is the host of the Ecomhunt Podcast. Can be followed on Twitter at @ArielBenSolomon