OpenAI introduces a new AI model named Sora 26, claiming it can create imaginative and realistic videos from user text prompts.
The company discussed that Sora can generate videos lasting up to 60 seconds. Sora also has the ability to produce scenes with different characters, each with specific kinds of motion, and different backgrounds.
OpenAI also clarified how their model stands out, stating what sets it apart is the length and accuracy it provides. With their take on text-to-video generative AI, Sora has the potential to impact personalized content creation.
But overall, Sora is still in development with areas to improve on. Specifically, understanding things like cause and effect.
OpenAI provided an example on their blog in the form of a created video. For instance, the scene starts with a person biting a cookie. The issue here is that the cookie did not appear to have a bite mark at all.
Concurrently, OpenAI is focusing on user safety. The company is planning to gather experts to test the latest model to focus on misinformation, hateful content, and bias. They are also building tools to detect misinformation to protect netizens.
Sora will initially be only available to cybersecurity professors, who are providing assessments on its risks. Selected visual artists, designers, and filmmakers also have access for feedback on how professionals in their industry could use it.
But as always, stay tuned for more developments in this story. We’ll be sure to let you know for any updates regarding Sora from OpenAI.
