Sora Filmmaker Tool, a new member of the OpenAi family

Just give him the text and he will make the movie you want in a minute. Sora, the newest member of the OpenAi AI family, unveiled on Thursday, adds video-making capabilities to the mix. This tool is named after the Japanese word Sora, which means “sky”.

Sora can produce one-minute videos according to the user's text commands, with high detail and very close to reality. In addition to making movies from text commands, Sora will be able to make movies based on an image or add new things to the elements of a movie.

“We've taught AI to understand and simulate the physical world in motion with the goal of training models that help people solve their problems,” OpenAi says on its blog.

Below you will see one of several early examples of the results of working with Sora, which was based on the following instructions: “A trailer featuring the adventures of a 30-year-old male astronaut wearing a red woolen knitted motorcycle helmet against a background of blue sky, salt desert and Vivid colors are seen. The style of the video should be cinematic and the film should be made on 35 mm size.

The company announced that it has provided access to Sora to several researchers and video creators. OpenAi experts also review this product for prohibited and sensitive content such as extreme violence, sexual content, hateful images, likeness of celebrities or other people's IP.

Currently, OpenAi only allows limited access to researchers, visual artists and filmmakers, though CEO Sam Altman responded to user requests on the network after the announcement by posting video clips he said were made by Sora. X replied. These videos have a watermark that indicates they were created by artificial intelligence.

OpenAI has not yet spoken about the number of videos and resources it used to train Sora; But the New York Times quoted the company as saying that the collection contains videos that are both publicly available and licensed from copyright owners.

Previously, OpenAi has been sued many times for copyright infringement in the training of its artificial intelligence tool. Despite the fact that companies like Google and Meta also announce the development of their own video models, OpenAi has been able to provide amazing examples of the result of its model's work. to present

nntt