Runway 联合创始人分享企业 AI 预算、人才与用量观察
过去几个月，我一直在世界各地旅行，与我们的客户和用户见面。我总体上的感受——我在我们的全体会议上与团队分享过——是现在是时候了。从现在时刻的几点观察：
- 企业已经准备好在 AI 上投入资金。三年前，媒体与娱乐和营销领域的 AI 预算来自实验资金。现在，它们来自主线预算，从其他优先事项中挤出资金。
- 你观看的大量电视、电影、节目和广告很可能包含 AI 生成的场景，或者完全由 AI 生成，而你并没有意识到。根据我们从客户那里看到的情况以及他们如何使用我们的工具，我们估计大约 20% 的产出已经包含生成内容。
- 招聘仍然是首要任务。公司迫切需要优秀的 AI 原生人才。那些能从当前前沿模型中榨取最大价值的人，而且同样重要的是，能教会他人做到同样的事情。
- 企业令牌消耗量在三个月内翻了一番多。那些在市场中获胜的公司正在通过最大化令牌使用量来实现领先。如果你没有扩大使用规模，其他人就会比你更快。
- 起点也比以往任何时候都要大。平均企业合同规模在不到一年内翻倍。不再只是小规模试点了。终于看到这一点，真是令人欣慰。
对照原文
For the last couple of months I've been traveling around the world to meet our customers and users. My general sentiment, which I shared with our team at our all hands, is that the time is now. A few observations from now o'clock: - Enterprises are ready to spend on AI. Three years ago AI budgets in M&E and marketing came out of experimental funds. Now they're coming from the main line budget taking from other priorities - There is a big chance that a lot of the TV, films, shows and ads you watch have AI generated scenes, or are fully AI gen, and you don't realize it. Based on what we're seeing from our customers and how they use our tools we estimate that around 20% of what's coming out already includes generated content - Hiring remains a top priority. Companies are desperate for great AI native talent. People who can get the most out of current frontier models and just as importantly, teach others to do the same - Enterprise token consumption more than doubled in three months. The companies winning their markets are tokenmaxxing their way there. If you're not scaling usage, others are moving faster than you. - The starting point is also larger than ever before. The average enterprise contract has doubled in less than a year. Not more small pilots anymore. Which is nice to finally see.