据IT之家报道，知名人工智能研究机构OpenAI近日推出了一款名为Sora的文本生成视频模型，该模型能够根据用户提供的提示词创造出逼真的视频内容。然而，这一技术创新的背后，也暴露出一些实际操作中的效率问题。据Reddit社区的网友反馈，使用Sora生成短短1分钟的视频，其渲染时间竟然超过1个小时，这一效率问题无疑为实际应用带来了挑战。
尽管Sora的技术演示引发了广泛关注，但研究人员目前仅展示了预先挑选的示例，并未向公众开放自定义提示词的功能。最长的公开演示视频仅有17秒，这使得外界对于Sora的实际性能和潜力仍抱有疑问。尽管视频生成的精度和真实感是人工智能领域的重大突破，但如此漫长的生成时间无疑限制了其在实时或大规模应用中的可行性。
OpenAI的Sora模型作为文本转视频技术的最新尝试，其在效率上的局限性也提醒了业界，技术进步的同时，如何平衡性能与实用性是一个不容忽视的问题。未来，OpenAI是否能优化算法，提升Sora的生成速度，将直接影响到这一创新技术在媒体、娱乐、教育等多个领域的应用前景。我们将持续关注OpenAI对此的回应和改进。
英语如下：
**News Title:** “OpenAI’s New Model Sora: Rendering 1 Minute of Video Takes Over 1 Hour, Sparks Efficiency Debate”
**Keywords:** OpenAI Sora, video generation, extended rendering time
**News Content:**
Title: OpenAI’s Model Sora: 1-Minute Video Creation Takes Over an Hour, Raises Industry Concerns
According to IT Home, renowned artificial intelligence research institute OpenAI has recently unveiled Sora, a text-to-video generation model that can produce realistic videos based on user-provided prompts. However, this technological innovation has exposed efficiency concerns in practical use. Users on the Reddit community have reported that rendering a mere 1-minute video with Sora takes more than an hour, posing a challenge for its practical application.
While Sora’s technical demonstrations have attracted widespread attention, researchers have so far only showcased pre-selected examples and have not opened up the custom prompt functionality to the public. The longest publicly demonstrated video is only 17 seconds, leaving outsiders with questions about Sora’s actual performance and potential. Despite being a significant breakthrough in the AI field for video generation accuracy and realism, the extended generation time significantly limits its feasibility in real-time or large-scale applications.
OpenAI’s Sora, as a recent attempt in text-to-video technology, highlights the need to balance performance with practicality as the industry progresses. The company’s ability to optimize algorithms and improve Sora’s generation speed will directly influence the adoption prospects of this innovative technology in sectors like media, entertainment, and education. We will continue to monitor OpenAI’s response and improvements in this regard.
【来源】https://www.ithome.com/0/751/364.htm
Views: 2