OpenAI’s new tool ‘Sora’ can turn text prompts into realistic videos

Sam Altman’s OpenAI is paving the way for imaginative artificial intelligence by launching a new tool called ‘Sora’, through which can can create very realistic-looking AI videos using text prompts and photos.

Sora, which means sky in Japanese, is a new model by OpenAI which can produce realistic videos of up to a minute after taking the instructions of a user on the style and subject of the clip.

The artificial intelligence company, in its recent blogpost, said that Sora also has the ability to generate realistic videos using still images or existing footage provided by the user for reference.

“We’re teaching AI to understand and simulate the physical world in motion, with the goal of training models that help people solve problems that require real-world interaction,” the blogpost by OpenAI reads.

One of the example prompts displayed by the company was “A movie trailer featuring the adventures of the 30-year-old space man wearing a red wool knitted motorcycle helmet, blue sky, salt desert, cinematic style, shot on 35mm film, vivid colors.”

The access to Sora was opened for a researchers and content curators to make sure that the new AI model is adhering to the security policies of OpenAI, which prohibit “extreme violence, sexual content, hateful imagery, celebrity likeness, or the IP of others,” the blogpost said.

After Sora’s launch, Sam Altman also took to X (formerly Twitter) and replied to a few users with videos based on their prompts. The videos contain a watermark which shows that the clip is make by AI.

Google and Meta have also tried to invest in AI technology like this, but OpenAI has taken it one step further with hyper-realistic videos based on just text prompts.

It must be noted that Sora is not available for public usage yet and OpenAI has released little information about how it actually works and how the model was created.

The new AI software is only available for red teamming right now, which will help the makers detect any flaw in its development and code. Meanwhile, OpenAI is also working on tools that can discern if a video was developed by AI.