Cheaper AI tokens are driving more demand, and that's Jensen Huang's best-case scenario
decoder ·
The article explains how falling AI token prices are driving increased demand, a phenomenon described as Jevons' paradox. Despite cheaper tokens, the rental costs for H100 GPUs remain steady or increase because lower prices unlock new applications and automation. This surge in volume outpaces the decline in per-unit costs, leading to higher overall compute requirements. A significant portion of this demand may come from agentic AI systems rather than human users, potentially inflating hardware needs. The market relies on continuous usage growth to offset falling token prices; if demand flattens, it could negatively impact chip makers, energy providers, and cloud companies. Recent stock drops following reports of lower-than-expected revenue at OpenAI highlight the sector's sensitivity. Ultimately, Jensen Huang's best-case scenario depends on sustained expansion in AI adoption to keep hardware scarcity and pricing high.
decoder