ChatGPT maker OpenAI shows off Sora, a text-to-video AI model

OpenAI has announced Sora which is a new text-to-video model, set to revolutionise content creation. The company showcased the model's capabilities with a host of videos generated using prompts.

Mehrotra A Neowin · Feb 16, 2024 00:22 EST with 19 comments

Earlier today, OpenAI announced Sora, a new text-to-video Artificial Intelligence (AI) model that is capable of turning written prompts into minute-long videos. The new model will go head-to-head with the likes of Google who recently announced Lumiere, a similar text-to-video AI model.

According to OpenAI's research paper, the new Sora model is trained on "text-conditional diffusion models jointly on videos and images of variable durations" to generate hyper-realistic videos from text-based prompts. The company noted that it is 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."

Introducing Sora, our text-to-video model.

Sora can create videos of up to 60 seconds featuring highly detailed scenes, complex camera motion, and multiple characters with vibrant emotions. https://t.co/7j2JN27M3W

Prompt: “Beautiful, snowy… pic.twitter.com/ruTEWn87vf
— OpenAI (@OpenAI) February 15, 2024

It also shared a bunch of clips on both the announcement post as well as the research paper showcasing the model, with company's CEO Sam Altman even taking prompt requests from users to show the model's capabilities.

we'd like to show you what sora can do, please reply with captions for videos you'd like to see and we'll start making some!
— Sam Altman (@sama) February 15, 2024

OpenAI further noted that Sora model is not available to everyone as the company is currently working with red teamers to test the limits of the model to ensure safety.

We’ll 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.
— OpenAI (@OpenAI) February 15, 2024

The company is also working with policy makers "to understand their concerns" surrounding the technology. While the model is not available to the public, OpenAI said that it will make the model available to select group of artists to get their feedback on the technology.

We are also granting access to a number of visual artists, designers, and filmmakers to gain feedback on how to advance the model to be most helpful for creative professionals.

We’re sharing our research progress early to start working with and getting feedback from people outside of OpenAI and to give the public a sense of what AI capabilities are on the horizon.

There is no word on when OpenAI will offer the new model to the public and also what would it cost to the end users. OpenAI has been leading the charge with its CEO Sam Altman rumoured to be looking for an investment to the tune of $7 Trillion for his endeavors.