OpenAI’s ’Sora’ can generate realistic videos with just a single prompt

Sora is an AI model that can generate realistic and imaginative videos up to a minute long from a user’s prompt. It builds on past research in DALL·E and GPT models using the recaptioning technique from DALL·E 3, which involves generating highly descriptive captions for the visual training data. In addition to being able to generate a video solely from text instructions, the model is able to take an existing still image and generate a video from it, animating the image’s contents with accuracy and attention to small detail. Although it is attracting praise from some AI researchers, it is also raising concerns about the risks of video deepfakes during global elections in 2024. Learn more ➤ Subscribe ➤ Get more from New Scientist: Official website: Facebook: Twitter: Instagram: LinkedIn: About New Scientist: New Scientist was founded in 1956 for “all those interested in scientific discovery and its social consequences”. Today our website, videos, newsletters, app, podcast and print magazine cover the world’s most important, exciting and entertaining science news as well as asking the big-picture questions about life, the universe, and what it means to be human. New Scientist