Artificial intelligence has learned to instantly create video from text: a powerful breakthrough

AI has opened up new horizons

Technology company OpenAI is launching a tool that uses artificial intelligence to instantly create video from text. The model from ChatGPT maker “simulates the physical world in motion” for up to a minute, based on user guidance on theme and style.

On Thursday, OpenAI unveiled a tool that can generate videos from text prompts, The Guardian writes.

The new model, called Sora after the Japanese word for “sky,” can create realistic footage up to a minute long, matching the user’s instructions in both subject matter and style. According to the company’s blog post, the model is also capable of creating video from a still image or augmenting existing footage with new material.

“We teach AI to understand and model the physical world in motion, with the goal of creating learning models that help people solve problems that require interaction in the real world,” the blog post says.

One video, included among several initial examples from the company, was based on the prompt: “Movie trailer about the adventures of a 30-year-old astronaut wearing a red wool knit motorcycle helmet, blue sky, salt desert, cinematic style, shot on 35mm film, vibrant colors”.

The company announced that it has made Sora available to several researchers and video creators.

Experts will “re-validate” the product—testing it against OpenAI’s terms of service, which prohibit “extreme violence, sexual content, hateful images, similarities to celebrities, or other people’s IP,” according to a company blog post.

The company provides limited access only to researchers, visual artists and filmmakers, although CEO Sam Altman responded to user queries on Twitter following the announcement with video clips he said were made by Sora. The videos have a watermark indicating they were made by artificial intelligence.

The company debuted still image generator Dall-E in 2021 and chatbot ChatGPT in November 2022, which quickly gained 100 million users. Other artificial intelligence companies have introduced video generation tools, although these models were only able to generate a few seconds of video footage that often had little relevance to their clues. The tech giants have said they are in the process of developing video generation tools, although they have not released them to the general public.

On Wednesday, the company announced an experiment to add deeper memory to ChatGPT so it can remember more of its users’ chats.

OpenAI did not disclose how much footage was used to train Sora or where the training videos might have come from, other than to report to the New York Times that the corpus contained videos that were both publicly available and licensed from copyright owners. The company has been sued repeatedly for alleged copyright infringement in training its generative artificial intelligence tools, which digest gigantic volumes of material scraped from the Internet and imitate the images or text contained in those data sets.