Why smarter AI models could drive up compute prices 10x

AI labs face a widening gap between 10x annual revenue growth and a constrained 3x increase in available compute capacity. This disparity forces labs to either expand margins, increase compute prices, or shift resources toward inference. Current trends indicate that rising compute costs, driven by high-value AI applications, are becoming the primary mechanism for balancing this demand. As models achieve human-level utility, their ability to monetize compute effectively creates a high barrier to entry, favoring labs that can optimize for efficiency. Physical bottlenecks—including fab capacity, EUV machine availability, and wafer allocation—suggest that this 3x scaling limit is difficult to surpass. Consequently, the industry is entering a regime where intelligence is increasingly scarce and expensive, reinforcing strong economies of scale and power concentration among the leading AI labs.