Answer extracted from the Proven Podcast — listen to the full episode below.
Can AI really transform healthcare access and clinical practice?
AI could meaningfully improve diagnostic accuracy, how doctors practice medicine, patient record management, and the ability to treat various conditions with more personalized drugs and treatments at lower costs. Yet the biggest impact of AI in healthcare so far has been in revenue management and negotiation between providers and insurers rather than direct clinical improvements.
The promise of artificial intelligence in healthcare sounds straightforward: better diagnostics, faster treatment decisions, smarter data management. But according to David Goldhill, founder of Sesame, the American healthcare system has a habit of deploying innovation in ways that don't directly benefit patients. AI is no exception to this pattern.
The disconnect is structural. An AI-driven vision could result in declining costs, improved data quality, and better patient outcomes — but only if the incentives in the system aligned with patient benefit rather than profit extraction. As Goldhill explains on the Proven Podcast, the U.S. healthcare system currently operates under a different logic: nobody in the system can afford to lower their prices and remain profitable, which creates a perverse incentive structure that no amount of AI sophistication can overcome.
"Nobody in the healthcare system can make more money, can be more profitable by lowering their prices."
David Goldhill — Founder and CEO of Sesame, a healthcare marketplace platform. Goldhill spent decades as an entertainment executive, running television operations at Universal Studios and the Game Show Network, building networks and television content, before applying his expertise to transforming healthcare through direct price transparency and consumer choice.
This observation cuts to the heart of why AI deployment in healthcare has prioritized administrative and financial optimization over clinical innovation. Revenue management and payer-provider negotiation directly impact the bottom line for hospitals, insurers, and health systems — and that's where AI investment has concentrated. Improving diagnostic accuracy or patient outcomes does not translate into higher profit margins if the entire payment system is designed to extract value from opacity rather than efficiency.
The real opportunity for AI in healthcare depends on a fundamental shift: aligning incentives so that lowering costs and improving clinical outcomes also improves profitability. Without that realignment, Goldhill's work with Sesame demonstrates how direct price transparency and cash-based healthcare can bypass these broken incentives entirely, creating space for genuine clinical and cost improvement.
Why clinical AI innovation lags behind financial AI deployment
The American healthcare system spends more than 40% of the world's total healthcare budget while serving only 3% of the world's population — a disparity rooted in how money flows through the system. When AI is deployed to optimize that flow rather than to eliminate waste, the math works against patient benefit. Administrative costs alone consume 10–15% of direct care spending, and AI has proven more profitable when used to navigate and manage those administrative layers than when used to shrink them.
For AI to truly transform healthcare access and practice, it would need to operate in an environment where efficiency generates profit. Today, that environment barely exists in America's insured and provider-negotiated care system. This is why the Proven Podcast episode with Goldhill explores alternative models that remove the financial incentives standing in the way of both AI innovation and cost reduction.
See also
How does Sesame create price transparency and competition in healthcare services compared to insurer-negotiated rates?
Sesame lists consultations, appointments, procedures, labs, and diagnostics at cash prices set by physicians, clinics, or hospitals directly on its platform, enabling direct price comparison and competition outside the traditional insurer negotiation system.
What is the actual cost of innovation in pharmaceutical development and how does competition affect pricing once drugs are established?
The first pill to get approval costs $9 million because it requires all the research, studies, and regulatory testing. The second pill costs a penny to produce, making early-stage drug development the real innovation cost, not manufacturing.
How much of healthcare spending goes directly to administrative costs of the payment system rather than actual medical care?
Administrative costs for the payment system account for somewhere between 10% and 15% directly of the cost of care, translating to thousands of dollars per person annually that never reach actual medical treatment.