OpenAI is a leader in consumer artificial intelligence, revolutionizing text and image generation thanks to DALL-E and ChatGPT.
Now, with his latest suggestion, Sora, The company wants to enter the text-to-video generation spaceand seeks to push the boundaries of artificial intelligence creativity even further.
Sora offers the opportunity to do this Produce videos up to a minute longcomplete with “multi-character” scenes, nuanced movement, and “detailed backgrounds,” all controlled by user instructions.
The results are surprisingly realistic and show the undeniable potential of artificial intelligence in visual storytelling.
In several demos provided by OpenAI, Sora easily brings prompts to life. From an oriental woman in the middle of the city to a scene with a hairy monster next to a melted candle that looks like something out of a Pixar film.
However, Sora is not without his faults. In some cases, small inconsistencies or unrealistic movements become apparent, highlighting areas that need perfection.
Sora is currently accessible to a select group of users whose role is to assess the potential risks and impacts. In addition, OpenAI works with visual artists, designers and filmmakers to collect feedback with the aim of improving the overall usefulness of the model for content creators.