Co-director of our AI lab and leader in healthcare innovation.

If healthcare were a film, we are watching it through a single, blurry frame. Each day, the human heart beats roughly 90,000 times—more than 32 million cardiac cycles a year. Yet, when assessing heart health, we often rely on just a single 10-second ECG reading during an annual checkup—capturing a mere 10-15 heartbeats out of those 32 million. It's like taking a single chord from Beethoven's Ninth Symphony and claiming to understand the majesty of the Ode to Joy. We're stuck in a snapshot world when we need to be watching the full movie—seeing the dynamic, continuous story of human health unfold frame by frame. This limited snapshot approach leaves critical signals of impending disease buried, allowing conditions to progress unchecked. Most human health data—genetic codes, lab results, imaging studies, even daily step counts or subtle shifts in speech—remains idle and unused. The consequences of this longstanding approach are as subtle as the faint hum of an intensive care unit at 3 a.m. and as catastrophic as the patient who never makes it home.

This narrow view isn't limited to cardiology. At first glance, America's healthcare crisis seems oddly paradoxical: The United States invests a staggering $4.3 trillion in healthcare annually—$12,318 per person—nearly triple what other developed nations spend for better outcomes. Japan achieves superior health results spending just $4,823 per person, while France ($5,376), consistently ranked among the world's best healthcare systems, along with Canada ($5,370) and the United Kingdom ($4,653), provide comprehensive care at less than half our cost. Despite this enormous investment, Americans are not living longer. Life expectancy stalls at around 76.4 years, nearly eight years behind Japan's 84.5. Between 250,000 and 400,000 Americans die yearly from preventable medical errors, making it the third leading cause of death after heart disease and cancer. Most shocking of all, $935 billion—25% of all healthcare spending—vanishes into a black hole of inefficiencies, administrative burdens, and preventable complications. This isn't just a broken system; it's a catastrophic failure that wastes both money and human lives in equal measure.
Every 24 hours, the medical community produces roughly 5,000 new clinical studies—a torrent of information so vast that a physician would need 18 months of non-stop reading just to digest a single day's research. Picture a doctor trying to drink from this fire hose of knowledge while simultaneously caring for patients, each with their own complex medical histories. Yet the true crisis lies deeper: hospitals and clinics use only 12% of available patient data in treatment decisions. Imagine walking into your financial advisor's office and learning they've ignored 88% of your portfolio—your investments, debts, assets, and market trends—before making recommendations about your financial future. The outcry would be immediate and justified. Yet in healthcare, where the stakes are measured not in lives, not dollars, we accept this massive blind spot as routine.

Consider a teenager in the Midwest who suffers from unexplained seizures and muscle weakness. She is shuttled from specialist to specialist, subject to round after round of inconclusive tests. On average, patients with rare conditions endure a diagnostic odyssey of seven to ten years, seeing up to eight experts and spending upwards of $100,000 out of pocket. Some 70% become clinically depressed along the way. These aren't just statistics—they're lives suspended in a system drowning in data but starving for insights.

Artificial Intelligence promises to rewrite this story, transforming medicine from a collection of snapshots into a continuous film of human health. Like the splitting of the atom revealed hidden energy within matter, AI unveils patterns in health data invisible to the human eye. Think of a small clinic in rural Mississippi, where a solitary nurse practitioner juggles patients with minimal specialist backup. AI tools, trained on countless patient histories and outcomes, could arm that practitioner with an early warning system. By integrating local epidemiological trends, subtle shifts in a patient's vital signs, and a dozen other variables invisible to human eyes, the system might predict a brewing diabetic crisis or early-stage heart failure. Instead of a rushed ambulance ride hours later, the nurse can intervene today, stabilizing the patient and preventing a downward spiral.

Another scene: a mother in Chicago cares for a child with perplexing neurological symptoms. Traditionally, she might drift through a labyrinth of inconclusive appointments. With AI at her physician's fingertips, algorithms comb through millions of clinical notes, lab findings, imaging studies, and research papers from around the world, matching the child's peculiar combination of symptoms to a rare genetic mutation. A diagnosis that once took a decade of frustration and heartbreak might now emerge in weeks.

But just as J. Robert Oppenheimer stood in the New Mexico desert and witnessed the dawn of the atomic age, we must grapple with the gravity of what we are unleashing. AI, like atomic energy, is morally neutral. Will it power a healthcare renaissance that spreads around the globe, offering predictive care to a farmer in rural Kenya and a retiree in rural Montana as easily as to a Wall Street banker? Or will it stratify healthcare even further, concentrating high-tech tools in wealthy urban hospitals while leaving rural and marginalized communities farther behind?
The stakes extend beyond hospital walls. Just as the atomic age's benefits and burdens weren't distributed equally, healthcare AI could either heal or deepen our system's existing fractures. To prevent this, we must establish rigorous safeguards: independent oversight ensuring AI aligns with patient welfare, collaborative leadership uniting clinicians and data scientists, and transparency in algorithmic decision-making. Disease recognizes no borders—neither should the solutions we create.

Already, researchers at MIT have shown that AI can detect signs of asymptomatic heart conditions in 95% of supposedly normal ECG readings, and even predict impending cardiac events with remarkable accuracy. The implications are profound: no longer must doctors wait for symptoms to scream their arrival; they can listen to the quiet murmur of risk long before it becomes a roar. Major healthcare systems are beginning to integrate AI-driven decision support into electronic health records, helping doctors identify patients at risk for common but deadly conditions like sepsis or stroke. In academic hospitals, machine learning assists radiologists in sifting through CT scans and MRIs, enhancing their ability to spot early tumors or subtle brain lesions.

Technology need not be the enemy of empathy. In fact, done right, AI can liberate doctors from administrative drudgery, giving them more time to talk to patients and families, to listen, to explain. This marriage of silicon and stethoscope promises to reshape medicine as fundamentally as the atomic age transformed warfare and energy. It transforms healthcare from a series of still photographs into a living, breathing film, where each frame tells part of a larger story. It might render the traditional tools no less important but will complement them with an invisible orchestra of algorithms humming quietly in the background, guiding decisions and averting catastrophes before they find their voice.

We can't unwrite the code any more than we could un-split the atom. But we can choose its purpose. As this technology moves from computer labs to clinic floors, we must ensure it amplifies rather than replaces the irreplaceable human elements of care—the hand on a shoulder, the attentive ear, the hard-won wisdom of experienced healers.

The old healthcare story reads like a tragedy of wasted resources and missed opportunities. The new story, still being written, shows AI offering us the pen. We now have the chance to redesign a system that works proactively rather than reactively, that values prevention as much as treatment, and that safeguards the human dignity at its core. As the heartbeat of change quickens, we must decide how to shape this future. The choices we make today will echo for decades, influencing who benefits from tomorrow's breakthroughs—and who gets left behind.