Is Anthropic's revenue the entirety of AI commercialization? Last night, Anthropic's investor day disclosed its revenue data for the third quarter, which was interpreted by the market as a slowdown in growth. It is difficult for ARR to reach $100 billion by the end of the year, and some people believe that there are major problems with commercializing AI. Can the revenue of cutting-edge model manufacturers represent the entirety of AI commercialization? In fact that was not the case. As early as early July, here is https://((x.com))/qinbufark/status/2074754779755295164? S=46&t=k6rimWSEbo2D2TXolYcM-A discussed that "the performance of future super large CSPs will be a more important indicator for measuring the commercialization of AI." The core logic is that the adoption of AI in the enterprise market is increasingly becoming a stage of segmented application models 1) Continue to use high-value tokens from closed source cutting-edge models in high-value scenarios; 2) In high-frequency, low complexity, and high volume task scenarios, enterprises are increasingly inclined towards high cost-effectiveness tokens. This is when enterprises adopt AI to enter the cost engineering stage. High frequency, medium low value scenarios will shift towards cheaper tokens, cheaper models, and cloud deployments that are closer to data. This will compress the unit economy of some closed source models, but not necessarily compress CSP; On the contrary, it may expand the total amount of inference and deposit commercial value in the cloud. The July tweet linked above provides a detailed explanation of the logic Here is also an example, Microsoft has disclosed in the second quarter that all additional orders came from non cutting-edge model companies. At the end of July, here is https://((x.com))/qinbank/status/2082657327799488557? S=46&t=k6rimWSEbo2D2TXolYcM-A has also been discussed, which fully demonstrates that the performance of super large cloud providers is more representative. Not only does the commercial revenue of cutting-edge closed source models mostly come from cloud providers, but now the revenue of non cutting-edge closed source models also falls on cloud providers. The commercial revenue from non cutting-edge closed source models here not only comes from open source models, but also from medium-sized models such as Microsoft's self-developed MAI model. Returning to the market, the tweet from last Friday afternoon on October 9th is https://((x.com))/qinbufark/status/2108458080510132423? S=46&t=k6rimWs Ebo2D2TXolYcM-A said, "The market will be very unstable in the near future, and the risk aversion, selection, and industry concerns before CPI will all lead to a slight wind and agitation, which can trigger adjustments. High Beta and momentum stocks are prone to ”， It seems that momentum stocks have fallen sharply in recent trading days. It was also mentioned in the tweet at that time that ultimately, the financial reports of several super large cloud providers at the end of the month were needed to give market confidence. I tend to believe that the financial performance of cloud vendors in the third quarter will greatly exceed expectations. Here are a few examples: 1) One is that the guidance given in the second quarter is very strong; 2) The craze for agents in the third quarter; 3) The adoption by enterprises accelerated in the third quarter. Looking at it from a different perspective: entering the earnings season with extremely pessimistic expectations is actually better than entering the earnings season with full confidence expectations. This article is sponsored by @ bitget_zh, "Bitget Buying US Stocks: Instant Entry, Smooth Trading