The demand for AI is rapidly outstripping the capacity of physical infrastructure. Data center growth is colliding with limitations in power grids, water access, and permitting, making these real-world resources the ultimate gatekeepers for the expansion of AI capabilities.

The AI buildout is unlikely to suffer a massive oversupply crash because it is constrained by real-world factors beyond chips: a lack of power, data centers, and even skilled trades like electricians. This acts as a natural governor, creating a longer, more durable investment cycle.

The total addressable market for AI is massive and not a concern. The real growth limiters are physical constraints like power grid capacity, permitting delays, and shortages of skilled labor and equipment. These "atoms and energy" problems will likely prevent the industry from building out compute as fast as forecasted.

The focus in AI has evolved from rapid software capability gains to the physical constraints of its adoption. The demand for compute power is expected to significantly outstrip supply, making infrastructure—not algorithms—the defining bottleneck for future growth.

Beyond chip packaging and memory, the next major constraint on AI growth could be the physical construction of data centers. Arm's CEO points to project delays, labor shortages, and local regulatory opposition as key headwinds that will throttle the expansion of compute infrastructure.

While demand for AI compute is massive, a potential overbuild by hyperscalers is naturally limited by real-world shortages of energy ("watts") and manufacturing capacity ("wafers"). These physical constraints may act as a governor on the market, preventing a classic tech over-investment bubble and bust cycle.

For companies like Anthropic, the primary obstacle to continued 10x year-over-year growth is no longer the total addressable market. Instead, it's physical limitations such as the availability of compute, data center capacity, and energy, which cannot scale as quickly as software demand.

While chip fabrication is complex, the most binding constraint for AI compute providers is physical infrastructure. The entire industry's growth is bottlenecked by the availability of powered data center buildings, a problem projected to persist for at least another 15-18 months.

Hyperscalers face a new economic reality where massive AI CapEx must be justified by durable revenue. This shifts their model from high-margin software to a more capital-intensive one, like railroads or oil, creating a timing-sensitive "matching problem" between spending and cash flow.