Uber's COO Andrew MacDonald stated the company is finding it increasingly difficult to justify its large expenditures on token-intensive AI operations, signaling a broader reckoning with compute costs across enterprise deployments. For healthcare AI vendors and consultants, this reflects a critical inflection point: the era of unlimited spending on computational resources without demonstrable clinical or financial returns is ending. Healthcare organizations pursuing AI implementation must now demand clear value metrics tied to clinical workflows, EHR integration efficiency, and measurable outcomes in value-based care arrangements. Vendors building bloated models will face pressure to optimize performance per token spent, creating opportunities for lean, purpose-built solutions that deliver ROI on specific clinical tasks rather than general-purpose capabilities.