Lily:Adam,I've been reading through this quarter's earnings, and one thing really caught my attention.
For the past two years, investors have mostly debated whether AI spending would ever generate an adequate return.
This quarter, the discussion felt different.
Microsoft reported Azure growth of 43% and said Microsoft 365 Copilot now has more than 30 million paid seats. The conversation wasn't about whether Copilot works—it was about enterprise adoption at scale.
Meta spent surprisingly little time talking about Llama itself. Instead, management focused on how AI is improving advertising performance while continuing to justify higher infrastructure investment.
Amazon made a similar point. Andy Jassy argued that demand for AWS AI services still exceeds available capacity, which is why the company continues to accelerate infrastructure investment.
What struck me is that none of them seemed interested in proving they had the best AI model.
They were all trying to prove something else:That AI is becoming a practical business tool,one that can drive revenue, strengthen customer relationships, and justify long-term capital investment.
It made me think of something you said during our conversation—that AI is ultimately a tool.
Perhaps this earnings season was the first time the market started valuing these companies on how effectively they use that tool, rather than on the tool itself.
何滟：Adam，我最近一直在看这一季度几家科技巨头的财报，有一个变化特别引起我的注意。
过去两年，市场一直在争论：这么大规模的AI投入，到底能不能带来足够的回报？
但这个季度，讨论似乎变了。
微软Azure增长43%，Microsoft 365 Copilot的付费用户已经超过3000万。大家讨论的已经不是Copilot到底有没有用，而是企业正在多大规模地采用它。
Meta也很少再强调Llama本身，而是更多讨论AI如何改善广告业务，以及为什么公司仍需要继续增加基础设施投资。
Amazon也是类似的逻辑。Andy Jassy表示，AWS的AI服务需求仍然超过现有供给能力，因此Amazon还在继续加大基础设施投入。
最让我感兴趣的是，这些公司似乎都不再急于证明自己拥有最好的AI模型。
它们正在证明另一件事：AI正在真正成为一种商业工具，能够创造收入、增强客户关系，也能够支撑长期资本投入。
这让我想到你之前说过的一句话：AI归根到底是一种工具。
也许这个财报季真正发生的变化是：市场开始关注的，不再只是AI这个工具本身，而是谁能够真正把这个工具用好。
Adam Seessel：AI is a tool, and it’s unclear which tool is best and will dominate. Maybe none will. Who knows? It’s murky and therefore investable.
What isn’t murky — what is very clear — is that the world’s going to need a lot more compute. Also, only 3 players can deliver that at scale. So that thesis is investable.
In general, I don’t pay much attention to quarterly earnings and the related chatter around them. They are like paving stones in a company’s long road — importanr, but not as important as the long road ahead.
亚当·西塞尔：AI是一种工具，但现在还不清楚哪一种工具最好、最终谁会占据主导。也许没有谁能真正一家独大。谁知道呢？现在还很模糊，而这种不确定性本身也意味着投资机会。
但有一件事并不模糊，反而非常清楚：未来全球需要的算力一定会大幅增加。而且，真正能够大规模提供这种算力的，目前只有三家主要玩家。所以，这个投资逻辑是清晰、可投资的。
总体来说，我并不会太关注季度财报，以及围绕财报产生的各种市场讨论。季度财报更像是一家公司漫长道路上的一块块铺路石。它们当然重要，但更重要的，是前面那条很长的路。
Lily:That really resonates with how we think at DVI.The question is not simply who has the best technology today, but which companies can turn technology into durable economic value.That means looking beyond a single model or quarter—at where a company sits in the AI value chain, and whether its advantages can ultimately translate into long-term growth and cash flow.Technology changes quickly. The real question is which value can compound over time.
何滟：这与DVI一直以来的思考非常一致。
真正的问题，不只是谁今天拥有最好的技术，而是哪些企业能够把技术真正转化为持续的经济价值。
因此，我们需要跳出单一模型、单个季度去看一家企业：它处在AI价值链的什么位置？它所拥有的优势，最终能不能转化为长期增长和现金流？
技术变化很快，真正重要的是：什么样的价值能够长期复利。
Adam Seessel是美国知名价值投资人、《Where the Money Is》作者，提出了广受关注的Value 3.0 投资理念。
2003年，Adam Seessel在纽约创立投资管理机构Gravity Capital Management，并兼首席投资官（Founder & Chief Investment Officer）。在创立Gravity之前，Adam曾先后任职于Sanford C. Bernstein、Baron Capital 和 Davis Advisors，是美国成长价值投资领域具有代表性的投资人之一。
释义：DVI是中创数经信息服务（上海）有限公司拥有知识产权的上市公司数字化价值投资评价体系和指数生态简称，2026年DVI的评价在数字化评价基础上，进一步延展“数字化+智能化价值投资”评价纬度与方法。
中创数经信息服务（上海）有限公司
地 址：上海市静安区康宁路288弄数通链谷园区7号楼联系人：何女士手 机：13801697923邮 箱：heyan@irongliang.com
