OpenAI，这个在人工智能领域享有盛名的机构，近日推出了一款创新的文本生成视频模型——Sora。这款模型能够根据用户提供的简单提示，创造出栩栩如生的视频内容，开启了文本到视频生成的新篇章。然而，尽管Sora的技术前景令人兴奋，但其实际应用还面临一些挑战。
据网友在Reddit社区的反馈，Sora在生成1分钟视频的过程中，需要耗费超过1小时的渲染时间。这一时间成本无疑对实时性和效率提出了考验，对于需要快速制作视频的用户来说，可能并不理想。目前，OpenAI仅展示了预先选定的示例，并未开放公众使用自定义提示词的功能，最长的公开演示视频仅有17秒。
尽管Sora的技术演示有限，但其在文本理解和视觉生成上的进步仍让人印象深刻。这一技术如果能克服效率瓶颈，无疑将为创意产业和媒体行业带来革命性的变化。OpenAI的这一尝试，再次证明了他们在推动AI技术边界上的决心和实力。我们将持续关注Sora的发展，期待它在未来能实现更快、更高效的视频生成，为用户带来更为便捷的创作体验。
英语如下：
News Title: “OpenAI’s New Model Sora: Rendering 1 Minute of Video Takes Over 1 Hour, Testing Patience in Speed”
Keywords: OpenAI Sora, slow video generation, limited demonstrations
News Content:
OpenAI, the renowned institution in the field of artificial intelligence, has recently unveiled its innovative text-to-video generation model called Sora. This model can create lifelike videos based on simple user prompts, marking a new era in text-to-video conversion. However, despite its exciting technological prospects, Sora encounters some practical challenges.
According to user feedback on Reddit, the process of rendering a 1-minute video with Sora takes over an hour. This time-consuming aspect raises questions about its real-time efficiency, which might not be suitable for users requiring swift video production. Currently, OpenAI has only showcased pre-selected examples and has not opened the feature for public use with custom prompts. The longest publicly demonstrated video clocks in at just 17 seconds.
Despite the limitations of Sora’s demonstrations, its advancements in text understanding and visual generation are remarkable. Should it overcome its efficiency hurdles, this technology could revolutionize the creative and media industries. OpenAI’s endeavor underscores their commitment and capability in pushing the boundaries of AI. We will keep an eye on Sora’s development, hoping for faster and more efficient video generation in the future, providing users with a more streamlined creative experience.
【来源】https://www.ithome.com/0/751/364.htm
Views: 1