Dr. Dave Chokshi stood at a podium in New York City on May 12 and made a provocative argument: artificial intelligence's greatest promise in healthcare may have nothing to do with discovering the next miracle cure.
The former New York City Health Commissioner, speaking at The New Wave of AI in Healthcare 2026 conference presented by the Windreich Department of Artificial Intelligence and Human Health at the Icahn School of Medicine at Mount Sinai and The New York Academy of Sciences, reframed how we should measure AI's success in medicine. Rather than celebrating its power to unlock new drug targets or spot patterns humans can't see, he asked a different question: what can AI help us actually deliver to the patients medicine still leaves behind?
This distinction cuts to the heart of a stubborn problem Dr. Chokshi calls the "discovery-delivery gap" or the "no-do gap"—the chasm between what science has already proven works and what patients actually receive. "We literally have medicines that are curative right now, or have near-perfect efficacy in preventing diseases right now that do not reach the patients who would most benefit from them," he said.
The examples he offered are stark. Curative hepatitis C antivirals have existed for more than a decade, capable of eliminating the virus before it causes liver cancer or requires transplantation, yet less than a third of diagnosed patients receive them. Lenacapavir, a twice-yearly injectable for HIV prevention, demonstrated 100% efficacy at preventing HIV in Phase 3 clinical trials—and yet remains out of reach for most who could benefit. For hypertension, a condition we have known how to manage for decades, half of all patients with high blood pressure remain uncontrolled. The barrier isn't scientific; it's systematic.
Dr. Chokshi's vision for AI centers on closing this gap through what he calls "case finding"—using artificial intelligence to help health systems identify people who may have an undiagnosed condition, qualify for proven interventions, or have fallen out of care before completing treatment. Rather than replacing clinical judgment, AI becomes infrastructure for surfacing the patients most likely to be missed and connecting them sooner to care already known to work.
But case finding is only the beginning. Dr. Chokshi emphasized what he called "navigation"—the complex, often exhausting work of moving patients from diagnosis through treatment, across scheduling, prior authorization, follow-ups, and completion of care. Healthcare systems lose people not because the science fails, but because the delivery system is fragmented, burdensome, and indifferent to patients' circumstances. This is where AI could become truly transformative, he argued: not as a substitute for care, but as infrastructure for follow-through.
The distinction matters ethically. If AI is deployed mainly to make already efficient systems more profitable or to benefit patients who already have access, it could widen existing gaps. But aimed at "the patients healthcare does not see"—those routinely missed in clinical rounds and care systems—AI becomes a tool for equity rather than efficiency.
"How do we direct AI, not just to the breakthroughs, but to the follow-throughs?" Dr. Chokshi asked. In that question lies a quieter, more persistent vision of healthcare transformation: not a race toward the next discovery, but finally delivering on the promises medicine has already made.