A new AI model has demonstrated the ability to generate highly realistic and coherent video from simple text prompts. The system, developed by researchers at a leading technology institute, represents a significant leap forward in generative AI for video. Unlike previous models that produced short, low-resolution clips, this new approach can create longer sequences with improved temporal consistency and visual detail. The technology has potential applications in film pre-visualization, game development, and educational content creation, though the researchers acknowledge concerns about potential misuse for creating deepfakes. They have stated they are implementing safeguards and are not releasing the full model publicly at this time. Read the full article at https://technologyreview.com/ai-video-generation-breakthrough.