Terms of Care, Part I: Imagining a Healthcare System That Actually Works

We live in the most information-rich moment in human history. So why does our healthcare system still feel like a maze of fax machines, missed calls, and vanishing records?

It’s April, 2025, and OpenAI’s Sam Altman just gave a TEDTalk introducing “agentic AI,” —a tool called Operator that can complete tasks on the internet on a user’s behalf. It would be fully capable of gathering available doctor appointments near me, checking to make sure they take my insurance, completing intake paperwork, and even sorting by preferences I might add like the provider’s gender or the color of the paint on the walls in the waiting room.

But, when I attend that appointment, the healthcare team won’t have access to my complete medical record —still scattered across multiple EHR systems, faxed documents, and handwritten notes. They likely won’t see the myriad of health and fitness data collected by my smartwatch. Nor will cross-checked data points about certain social determinants of health— ecological surveys about the environments I live and work in, my relative housing and food security, or loneliness— enter into their decision making.

This is an era of unprecedented access to information, and industries all across the economy are bracing for impacts of AI. Healthcare is no different. We’ve all seen the attention grabbing headlines about AI outperforming doctors, even when doctors were assisted by AI. Most experts agree: AI is still a long way from replacing physicians. I’d go further. Given the complexity of human illness, the dynamic nature of health and wellness, and the massive gaps in public health literacy and access to care, we need doctors now more than ever.

But healthcare is changing, and I think that limiting these tools or our thinking about them to “personal assistants” misses the potential of a more accessible, personalized, interconnected healthcare system. A system that’s easier to navigate for providers and patients and administrators.

Two Futures

Imagine a young woman, Rosie, waking up with a bad headache. She used to get headaches in middle-school but hasn’t had any problems for years. In fact, she hasn’t been to any kind of doctor in a couple years.

She downloads an app on her smartphone, then enters her name, date of birth, and Social Security number. Within minutes, the app pulls her medical history, prescriptions, and clinical notes. It prompts her: “Would you like to sync Apple Health & Fitness data?” She agrees, and data from her smartwatch and iPhone are loaded into the app too. Everything is organized into widgets, with her current information displayed in the header.

But right now, Rosie just wants help with her headache. So, she navigates to the apps integrated AI assistant:

“Hello, I woke up with a bad headache this morning, I took 2 ibuprofen but they aren’t helping. Should I see a doctor?”

The app asks her a short series of follow-up questions about the onset, severity, and symptoms of her headache, which after evaluation, doesn’t appear to require urgent care. It gives her a reassuring message, noting that despite her concerns over the IB not helping and her medical history with headaches, this headache might be due to recent changes in her sleep pattern detected by her watch.

But, the app also gathers that her most recent appointment was 2.5 years ago, and asks:

“Would you like me to find available appointments nearby with a primary care physician and/or a sleep specialist?”

Rosie agrees. She’s been busy starting a new job in a new city and hadn’t made time to find a doctor. The app walks her through insurance verification, asks a few preference questions, and pulls up available providers nearby. Rosie books a primary care visit for next week, makes herself a big glass of water, eats breakfast, and—slowly—the headache fades.

But not everyone wants to wear a smartwatch, or talk to an AI health assistant through an app.

Let’s consider Roy, an 83-year-old retired farmer living just over an hour away from the nearest hospital.

Roy feels tightness in his chest one morning over coffee. His wife, Laura, doesn’t hesitate. She calls 911. EMTs arrive and load Roy into the ambulance. Laura provides his medical history, and the EMTs pull up Roy’s record through the same system Rosie used—an integrated Electronic Health System (EHS).

Roy’s condition worsens during transport. He receives nitroglycerin. One of the EMTs writes an update directly into his chart. That update is visible to the ER team in real time. The anesthesiologist spots Roy’s smoking history and prepares accordingly. Roy is taken straight into bypass surgery. His life is saved.

He spends two days in the hospital, then goes home with a new prescription. He is not billed—for the ambulance, the surgery, the medications, or the hospital stay. Roy is able to login to the EHS on his home computer and read the team’s chart notes to follow everything that happened to him while he was admitted, plus see organized dietary advice he went over with a social worker. Roy and Laura send thank-you cards to the EMTs and the care team at the hospital, and they are able to attend their grandkids graduation a month later.

These scenarios are fiction, but they shouldn’t sound impossible.

A comprehensive and cohesive healthcare system in America could look like this… it should work like this.

Many EHR vendors are introducing features like AI scribes into medical practices. Google has active projects with Covered California attempting to streamline insurance eligibility for Californians using AI. Under the 21^st Century Cures Act, the Office of the National Coordinator for Health Information Technology (ONC) organized and published the Trusted Exchange Framework and Common Agreement (TEFCA), which aims to eliminate technological and administrative barriers sharing health records between healthcare providers, patients, public health agencies, and payers.

The problem is that without a cohesive vision, these efforts may never change the qualitative experience of healthcare for Americans who would benefit.

America’s healthcare system is hardly a system at all. We have the most fractured, most expensive, and most inconsistent system in the developed world. We have the private hospital system, the pharmaceutical industry, the scientific instruments industry, the insurance industry, the home-care industry, the care home industry, private practice physical therapy, chiropractors, social workers, care coordinators, administrative specialists, and debt-collectors…

We want a healthcare system that provides timely, effective, and patient-centered care without fears of paperwork and bankruptcy. Over the next three articles, I’ll get into why the American healthcare system operates like it does, what it needs to achieve this vision for the future, and what I consider to be the most reasonable path to get there.