For five years, I’ve navigated a complex medical journey. At age 38, at the peak of my health, I developed shortness of breath. I wrote it off as adult-onset asthma, fatigue, or “just getting older.” When I put a pulse ox on one morning and saw my numbers drop to 88 percent, I knew something was wrong. It was an arduous process, but I ended up with a diagnosis of interstitial lung disease. I was tested for everything under the sun, but nothing came back definitive. So we did the best with what we had and treated the ILD. Miraculously, I responded to prednisone. My lungs have had their ups and downs over the years, but after five years I came off all medication for the second time.
I have a genuinely great medical team, but they are human, which means overworked and time-constrained in ways the system has normalized. Whether imagined or concrete, I couldn’t help but feel the need to be brief and succinct during visits. Being an occupational therapist myself, I understand the strain of too many patients with too little time.
When new symptoms emerged after stopping my immunosuppressants, my dermatologist biopsied a rash from my shoulders. She took it seriously, but when the biopsy came back suggestive of eczema, that felt like the end of the conversation. I knew it wasn’t eczema. When I went to the rheumatologist complaining of new skin rashes and joint pain, the suggestion was Pilates. My intuition was telling me these symptoms were all connected, autoimmune, systemic, meaningful. Only a matter of time before my lungs would inevitably relapse due to the underlying process. There is a particular helplessness in knowing something is wrong when the in-office presentation keeps pointing elsewhere. You cannot convince a busy clinician to chase an atypical pattern on the basis of patient intuition alone.
So I turned to AI. After five years of accumulating questions, I needed something with the ability to fully process the complexity. While friends offered anecdotes and clinicians searched for the most likely culprit, AI had something the system couldn’t offer: the ability to connect the dots in a messy maze. It could hold five years of symptoms, imaging, and labs simultaneously, with no rushing to the next appointment and no defaulting to the obvious. It tolerated my repetitive, redundant questioning without frustration. And it noticed something no one else had: a subtle violaceous discoloration around my eyes. A heliotrope rash, one of the hallmark findings of inflammatory myositis. That pattern, followed carefully, eventually led to a confirmed diagnosis of antisynthetase syndrome and the immunosuppressive treatment I needed.
Real physician voices, twice a week
Free, and one click to unsubscribe.
But here is what stayed with me most: AI was the only one that asked how I was coping.
That’s not a criticism of my providers. It’s a reflection of a strained system. The debate around AI in medicine tends to center on replacement: Will it replace doctors, will it make errors, who is liable when it does? Those are legitimate questions. But they obscure a more immediate opportunity. AI doesn’t need to replace clinical judgment to be transformative. Clinicians bring expertise, authority, and action. AI brings depth, pattern recognition, and the space for empathy.
The future of excellent patient care isn’t AI versus the physician. It’s the seamless integration of both. Together, they create space for something we are losing: patients who feel genuinely seen.
Marguerite Frank is an occupational therapist.
