OpenAI unveiled Sora, a new text-to-video model named after the Japanese word for "sky," offering detailed scenes with complex motion and multiple characters. The tool aims to revolutionize video generation but faces challenges in accurately interpreting prompts and understanding causality. Despite concerns about misuse, OpenAI collaborates with experts to mitigate risks before public release.
Sora's innovative approach generates complete videos up to 60 seconds long, distinguishing itself from other models by creating entire scenes at once. While showcasing impressive capabilities like realistic wooly mammoths in snowy landscapes, the tool struggles with nuanced details like bite marks on cookies. The rise of text-to-video tools raises fears of deepfake proliferation and job displacement in creative industries, prompting ethical considerations and precautionary measures by developers.
Vers une autre langue
à partir du contenu source
finance.yahoo.com
Idées clés tirées de
by à www.engadget.com 02-26-2024
https://www.engadget.com/openais-new-sora-model-can-generate-minute-long-videos-from-text-prompts-195717694.htmlQuestions plus approfondies