The doctor will see you (and your algorithm) now

Artificial intelligence (AI) is changing healthcare from the inside out. Just two years ago, AI in healthcare sounded like a futuristic experiment confined to pilot programs and research labs. Now, in 2025, it’s woven into the daily workflow of most physicians and health systems. The conversation has shifted from if the technology should work alongside clinicians to how.

Dr Phil: “For me, AI isn’t about replacing the physician’s judgment; it’s about giving us back the time and focus to truly see our patients again.”

The great acceleration in AI adoption

In February 2025, the American Medical Association (AMA) reported that two‑thirds of physicians (66%) use AI in some aspect of their practice—a staggering 78% jump from 2023. That’s an extraordinary rate of change for an industry that seldom moves quickly.

Much of this enthusiastic adoption is coming from practical use cases that free clinicians from routine, time-consuming tasks:

Clinical documentation: Automating billing codes, visit notes, discharge instructions, and care plans

Translation and accessibility tools: Breaking language barriers with real‑time voice transcription and translation

Assistive diagnosis: Surfacing findings from imaging and lab data to help strengthen physician decision‑making

Over half of doctors (57%) identified “reducing administrative burdens” as the leading promise of AI. This makes sense. Administrative overload has been a major driver of burnout and AI automation is, finally, offering some relief.

Juan: “From a technology perspective, what excites me most is when AI actually becomes invisible—a quiet partner that integrates seamlessly with clinical workflow instead of adding more clicks.”

The new landscape of clinical AI

Today, the integration of AI surpasses paperwork. Imaging and diagnostics are among the most visible successes. Radiologists and ophthalmologists routinely use AI to detect cancers and retinal changes that might otherwise go unnoticed. However, NPR reported a cautionary finding: gastroenterologists who grew accustomed to AI‑assistance became 20% less accurate once the system was turned off.

Dr Phil: “That study hit home for a lot of us. It’s a reminder that AI should sharpen our intuition, not dull it. The tools are incredible, but we still have to keep thinking.”

Predictive analytics are also maturing quickly. Systems now flag sepsis and readmission risks in real time, prompting earlier interventions. Remote‑monitoring systems continually analyze chronic‑disease markers so doctors can proactively adjust treatments before problems escalate.

Meanwhile, surgical and pharmaceutical innovations show the next frontier. Johnson & Johnson’s Polyphonic ecosystem uses computer vision in operating rooms to measure performance and guide improvement. Its VirtuGuide™ platform blends imaging and navigation capabilities for more precise procedures. Meanwhile, its AI‑driven biomarker testing is advancing personalized medicine and faster clinical trial recruitment.

Juan: “These examples show how data and design, when done right, can expand what’s possible in patient care. But the technology has to stay accountable to the clinician’s experience.”

A double‑edged transformation

As AI takes hold, new ethical questions arise. The Association of American Medical Colleges recently cautioned that efficiency may come at human cost. Physicians freed from paperwork may choose to see more patients rather than spend more meaningful time at each appointment. If left unchecked, AI could improve output at the cost of empathetic care.

Dr Phil: “The best outcomes come when I can really listen. Sometimes, what heals isn’t in the chart at all. AI can’t replace that conversation.”

The New Yorker echoed this tension, noting that medical schools now train students to think with AI but also beyond it. One student admitted, “I’m worried these tools will erode my ability to make an independent diagnosis.” That highlights the balance medical education must strike: harnessing computational insights while protecting clinical intuition.

Education and the “explainable AI” movement

At Harvard Medical School, researchers are testing Dr CaBot, an AI system that shows its reasoning process out loud. It is designed to teach students how to think rather than what to conclude. Built on OpenAI’s newest reasoning model, Dr CaBot can sift through millions of medical abstracts while describing its logical reasoning every step of the way.

Juan: “Explainable AI may be one of the biggest trust builders we have. Transparency will decide which tools clinicians actually keep using.”

Building trust in the AI system

An AMA study found that while 35% of physicians now feel more enthusiastic than concerned about AI, they still demand safeguards: privacy, seamless workflow integration, feedback loops, proper training, and stronger regulatory oversight. Nearly half of physicians ranked increased oversight as their top requirement for trust in AI.

Dr Phil: “Doctors aren’t anti‑tech; we just need to know the system has our patients’ backs. Trust is built when AI makes care safer—not faster.”

The next phase: Human + machine medicine

If 2024 was the year of experimentation, 2025 is the year of trust and thoughtful adoption. The most forward‑thinking physicians now talk about AI as a colleague. It handles what machines do best—recognizing patterns and automating repetition—to enable clinicians to get back to what they do best: judgment, creativity, and empathy.

Juan: “The future of healthcare isn’t man versus machine. It’s man amplified by machine—an alliance that lets both do what they do best.”

AI’s rapid adoption in healthcare highlights both its promise and its peril. It can save time, detect disease earlier, and unlock hidden insights. But it can also create dependency and distance if not utilized carefully. The shared mission for technologists and clinicians alike needs to remain ensuring that AI brings us closer to the heart of medicine—not further from it.

Dr Phil: “Technology should make medicine more human, not less.”