咱们今天从这两天爆火的AI视频生成聊起，由于相关论文提供的细节内容还不够丰富，咱们暂时还无法对技术做特别深入的评论。 不过可以先说说大致的原理，以及外行普遍的误解。咱们首先要说，这个SORA模型在视频生成方面取得的突破是值得肯定的。它的核心思想在于对视频进 行高度的抽象和压缩，把视频映射到一个抽象的空间。 我这里给大家打一个简单的比方，大家应该都看过皮影戏，你可以把 SORA理解为一个高级点的皮影戏，它有一些抽象的基本的视频元素，相当于木偶，然 后用这些元素拼接成完整视频。 因为每个元素都是来自真实世界的视频，所以拼接出来的整体，至少"打眼一看"还是非常有真实感。然而，SORA和大家想象的真正的"理解物理世界"，然 后基于这个理解仿真出一个虚拟的世界，还有不小的距离。多大呢？就是你用手机拍视频和电影制作之间的距离。 然后咱们再客观的讨论一下这个技术的实际影响。可能最重大的影响：股市。 这两天简单翻了一下网上各路外行的评论，尤其是那些不懂技术但是有流量的所谓大V，我觉得他们的想象力真的是非常令人佩服。 因此不用说取代实拍电影，就算是取代动画电影，以目前已知的技术思路，也是绝无可能性。注意不是还有多大差距，是绝对 ...

Core Insights - The SORA model has made significant breakthroughs in video generation by abstracting and compressing video into an abstract space, akin to a sophisticated shadow play where basic video elements are combined to create a complete video [3][25][71] - There is a considerable gap between the current capabilities of AI in understanding the physical world and the expectations of generating realistic videos, comparable to the difference between smartphone videos and film production [3][25][71] - The stock market is likely to be significantly impacted by the hype surrounding this technology, as public perception often overestimates the capabilities of AI [3][25][71] Industry Application - The industry is still in the early stages of exploring practical applications for this technology, with many discussions focused on its potential to disrupt the film industry, although the feasibility of AI-generated films remains questionable [5][28][51] - Current AI technologies struggle with concepts like acceleration and cannot produce precise physical motion trajectories, making it impossible to replace live-action or even animated films with AI-generated content [7][30][76] - There is a misconception among the public that if AI can perform complex tasks, it can also handle simpler ones, which is not the case; AI excels in processing large amounts of simple data but struggles with tasks requiring nuanced understanding [8][31][77] Market Dynamics - The volatility of the stock market encourages speculative investments in technologies that capture public imagination, leading to cycles of hype and disillusionment [11][34][57] - Companies that focus on clear business models and practical applications of AI, such as Huawei, tend to avoid the speculative hype and instead prioritize sustainable revenue generation [13][36][61] - The distinction between companies like OpenAI, which rely on speculative funding, and those like Huawei, which have clear revenue streams, highlights different approaches to AI development and commercialization [13][38][63]