How To Try OpenAI's Sora — The New Text-To-Video Model

On Thursday, ChatGPT-parent, OpenAI dropped its latest AI maverick Sora, a new model designed to generate high-quality videos from text prompts. So, naturally, when the tech world got the news, they simply termed the result as "flipping spectacular." 

OpenAI's text-to-video model can create videos of up to one minute – from snowy Tokyo city, giant wooly mammoths to a highly detailed Sci-Fi movie trailer, with simple prompts.

While these capabilities have generated quite a bit of interest among tech enthusiasts, some people on social media are also highly concerned about Sora making their video production jobs obsolete. 

See Also: How To Use ChatGPT On Mobile Like A Pro

People are also worried about the technology being used for spreading misinformation, disinformation, and hateful content, especially in the wake of the elections. 

For the people concerned, the Sam Altman-led company has clarified that they will be taking "several important safety steps ahead of making Sora available in OpenAI's products."

"We are working with red teamers — domain experts in areas like misinformation, hateful content, and bias — who are adversarially testing the model," the ChatGPT-maker said on Wednesday.

"For example, once in an OpenAI product, our text classifier will check and reject text input prompts that are in violation of our usage policies, like those that request extreme violence, sexual content, hateful imagery, celebrity likeness, or the IP of others," the company said. 

Adding, "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."

How To Try OpenAI Sora Right Now

Another concern has been echoing throughout the social media channels since OpenAI shared the amazing videos generated by Sora. "How can I try it?"

The short answer to this question is: You cannot. The long answer we have explained below.

Although Sora was officially revealed, the platform has undergone red-teaming. The company has provided access to a specific cohort, including visual artists, designers, and filmmakers. The idea is to gather insights on improving Sora's utility for creative practitioners. 

At this point, no specific schedule for a broad release has been disclosed by OpenAI. Therefore, unless you belong to the red-teaming group or the creative testing cohort, you must be patient and manage the available demonstrations.

Having said that, it doesn't mean you should completely give up. Try sending the video prompt to Altman directly on X, formerly Twitter, as he has been generating videos via Sora based on the internet’s requests. 

Here are some examples: 

It is important to note that OpenAI has stated that Sora’s current model exhibits certain weaknesses. 

It could encounter difficulties accurately simulating the intricate physics of a scene and might not grasp particular instances of cause and effect. "For example, a person might take a bite out of a cookie, but afterward, the cookie may not have a bite mark.”

Moreover, the model might misinterpret spatial details within a prompt, such as confusing left and right orientations. It could also encounter challenges in providing precise descriptions of events unfolding over time, such as tracking a specific camera trajectory.

Check out more of Benzinga's Consumer Tech coverage by following this link.

Read Next: OpenAI Sets Sights On Something Its Backer Microsoft Failed To Do: Beat Google At Web Search