From text to video: OpenAI unveils revolutionary neural network Sora
OpenAI introduced an AI model that generates video from text (photo: Unsplash)
OpenAI, the company that created ChatGPT, introduced a new neural network called SORA. This system can create realistic videos based on text descriptions or images, writes the official website of the OpenAI.
It is stated that compared to existing analogs, Sora is a real breakthrough in the field of artificial intelligence.
What is known about the new neural network
Sora is capable of creating complex videos with a high degree of detail, containing many dynamic objects and having an understanding of their physical properties. Generated videos can be up to one minute long.
.png)
Screenshot
The developers note that the current model can accurately model the physics of complex scenes and incorrectly interpret certain cause-and-effect relationships.
.png)
Screenshot
Following the announcement, Sam Altman, head of OpenAI, took to social media to demonstrate Sora's work, generating videos in response to requests in the comments.
The access to the neural network is currently limited to only a certain group of testers, as well as some illustrators, designers and cinematographers, who have to explore potential areas of application of Sora.
More examples of Sora videos
.png)
.png)
.png)
.png)
Screenshot
We earlier reported that a free analog of ChatGPT, which can work without the Internet, had been released.
