OpenAI reveals Sora, its new revolutionary AI-powered video generator!
The AI race continues and the new tool by OpenAI has finally been revealed: Sora can create videos based on input text within moments, and it's truly fascinating. Read on to find out more details about Sora.
What is Sora and how does it work?
OpenAI, the creative force behind ChatGPT and the DALL-E image generator, has introduced a new groundbreaking tool named "Sora." This innovative platform can generate realistic one-minute videos based on simple text input, representing a significant stride in artificial intelligence. During its experimental phase, access to Sora will be limited to a select group of creators.
Drawing inspiration from prior research on DALL-E and GPT programs, Sora is currently undergoing testing. Despite being in the experimental stage, OpenAI has showcased the tool's capabilities by presenting videos and their corresponding textual inputs. According to OpenAI, Sora can generate high-quality videos up to one minute in length while adhering to the user's input.
OpenAI calls its new system Sora, after the Japanese word for sky. According to Oren Etzioni, a professor at the University of Washington who specializes in AI, the team behind their technology chose this name because it "evokes the idea of unlimited creative potential."
How to use Sora
The versatility of Sora is evident in its ability to create intricate scenes featuring multiple characters, specific movements, and precise details. Additionally, Sora can transform a still image into a dynamic video and extend the duration of existing videos.
CEO Sam Altman has tweeted: "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". Users tweeted their ideas for the videos, and Altman shared some based on his Twitter followers. Notable examples include videos of dogs playing in snowy mountains and an imaginary creature soaring in front of a captivating sunset, with a hamster in sportswear on its back. The following is one such video, generated by Sora from this description suggested by one of the users: "A cooking session with instructions for preparing homemade gnocchis by a grandmother influencer and set in a rustic Tuscan kitchen with cinematic lighting."
OpenAI envisions Sora as a foundational tool for programs that comprehend and simulate the real world, aspiring to contribute significantly to the development of General Artificial Intelligence (AGI). It should be mentioned that Sora is still under development and has some limitations that make it generate illogical videos, in terms of physics principles or cause-and-effect relationships. For example, the system created a video where someone ate a cookie, but it never got smaller. Moreover, the tool's imitation of the camera shot can be inconsistent from one moment to the next. Sora also seems to have trouble rendering human hands and sometimes spontaneously generates new characters in random places, as if they magically appear. The developers are working on all these details.
When will Sora be available?
As mentioned, the AI model is still in the testing phase and is currently only accessible to a select group of OpenAI researchers. The company also stated that it is collaborating with experts to evaluate Sora's potential for misinformation, offensive content, and lack of fairness. OpenAI has not announced an official launch date, but all indications are that it will be integrated into their product offering once they take the necessary steps to ensure user security.
While many have shown enthusiasm for all the possibilities this type of technology offers, others fear it will put them out of a job. While Sora could speed up projects for experienced filmmakers, it is certain that it could completely replace less experienced digital artists.