OpenAI’s New Video Model Sora Sparks Excitement in AI Industry
People’s Daily Online, Beijing, February 26 (Jiao Lei) – OpenAI recently released its new video model called Sora, which has quickly captured the attention of industry experts and analysts. This innovative product is expected to drive more investment in the development of multi-modal large models and push forward the advancement of artificial general intelligence (breaking latest news).
The video field has long been seen as a promising sector for AI applications, and Sora’s release comes after the success of ChatGPT in generating graphics and text. Sora meets the industry’s demand for multi-modal large models, which can process various types of information such as text, pictures, audio, and video simultaneously.
In a research report by Huafu Securities, it was highlighted that multi-modal large models offer a more integrated approach to processing information, making them more aligned with how humans interact with the world. This advancement is expected to lead AI towards breaking latest news by improving human-computer interactions and enabling a wider range of tasks to be performed intelligently.
The State-owned Assets Supervision and Administration Commission recently emphasized the importance of AI in industry renewal, suggesting the need for AI+ special actions to empower key industries. The promotion of industrial multi-modal high-quality data sets and the development of a large model empowered industrial ecosystem are key strategies moving forward.
SPDB International analysis indicates that Sora’s impact on the content creation industry will be significant, particularly in short videos, advertising, and interactive entertainment. While short videos will benefit from improved quality and efficiency, long videos will require further enhancements in accuracy and editability to reach their full potential.
IDC China Research Director Lu Yanxia predicts that Sora will first make its mark in short videos, advertising, and media, revolutionizing the production process and user experience. However, the demand for computing power in AI video applications is expected to increase exponentially, posing challenges in availability and cost.
Furthermore, the security implications of using multi-modal large models for video generation are significant. Ensuring the compliance and safety of image and video data, as well as evaluating the difference between generated content and reality, will be crucial in maintaining personal and corporate security.
Looking ahead, dozens of large-scale Vincent video models are expected to emerge by 2023, with global user numbers surpassing one million. While most models are still in the early stages, start-up companies and tech teams are already showcasing their capabilities in this space. The future of AI video models holds immense potential, with advancements in computing power and security measures set to drive innovation in the industry.