A new AI system developed by researchers at Stanford University demonstrates the ability to generate realistic and coherent video from simple text prompts. The model, named VideoGen, uses a novel diffusion architecture that builds upon recent advances in image generation. Early tests show it can create short clips of animals, landscapes, and basic human actions with surprising temporal consistency. The researchers emphasize the work is a proof-of-concept and that significant challenges remain in generating longer, more complex narratives and ensuring the technology is not misused. The full details of the model and its training data are available in the research paper published in Nature. Read the full article at https://technologyreview.com/2024/05/15/videogen-ai-text-to-video.