The Horrible Economics of AI Are Starting to Come Crashing Down

AI firms are facing rising compute and data-center costs, putting pressure on prices and usage limits.

Setbacks in data-center buildouts are making access tighter, with more pay-per-token billing and stricter controls.

Experts warn current revenues may be far below what’s needed to fund AI infrastructure at scale.

The result could be the end of cheap AI, raising costs for businesses and consumers.