article5 min read•July 6, 2025

Dr. ChatGPT will see you now...

It all began on **Friday, June 6th**, with what felt like simple fatigue. Suddenly my eyes were heavy and tired, and I chalked it up to a long work week, or maybe some screen overuse, or recovering from the viral throat infection a week prior.

It all began on **Friday, June 6th**, with what felt like simple fatigue. Suddenly my eyes were heavy and tired, and I chalked it up to a long work week, or maybe some screen overuse, or recovering from the viral throat infection a week prior.

But by **Saturday morning at 6:00 am**, I was in extreme discomfort—pain, redness, sensitivity to light, and a gritty feeling in one eye. I was suddenly thrust into a spiral of not just physical discomfort, but emotional stress and confusion.

What kind of doctor do I need? Should I go to urgent care, the ER, or wait to speak to my primary care physician on Monday? Should I find an ophthalmologist or just start with an optometrist?

Here’s the problem: the healthcare system isn’t built to help you triage your own symptoms.

My insurance provider’s website lets me search for in-network providers, but it doesn’t tell me what kind of doctor to see for sudden eye pain on a weekend.

Caught in this maze, I opted for the fastest route—a telehealth visit at 6:30 am Saturday morning.

The physician prescribed an antibiotic eye drop (Polymyxin B with Trimethoprim) along with oral antibiotics, and I hoped for the best. But nothing changed.

My days consisted of waking up with severe eye pain early mornings (between 3am - 5am) which was correlated with increased blood glucose, extreme photosensitivity and dry eye symptoms with pain throughout the day. All I had to manage the pain were a combination of hot and cold eye compresses.

Over the next several days, my symptoms fluctuated, my blood glucose began to spike inexplicably (hovering above 300 mg/dl at times), and I grew more anxious. I had dropped 3lbs in the last 3 weeks.

I decided to meet my primary care physician on Wednesday, June 9th. My eye looked worse, not better. The uncertainty became unbearable. I was asked to continue the antibiotics for 3 more days, switch to Ofloxacin for 2 days if things didn’t improve and finally see an Ophthalmologist if the Ofloxacin didn’t clear things out.

Needless to say, things didn’t improve. The nagging question in my mind was –

Why I am being prescribed antibiotics when I am not showing any visible symptoms of bacterial conjunctivitis. What if it’s something else?

On Friday, June 11th, tired of the aggravating symptoms, lack of answers and lack of trust in the care regimen, I turned to **ChatGPT**.

What started as a simple query turned into a rich, evolving conversation with what felt like a knowledgeable and deeply attentive medical assistant.

I uploaded images of my eye, described my symptoms in real-time, and shared detailed context about my medical history.

I didn’t expect much. But I was stunned by the results.
Initial Analysis spit out by ChatGPT. Sorry for the grotesque image
Initial Analysis spit out by ChatGPT. Sorry for the grotesque image
Illustration
Illustration
Illustration

ChatGPT didn’t just react to symptoms. It listened. It remembered. It tracked progression. It synthesized my eye discomfort, recent viral illness, and sudden glucose elevation into a **plausible clinical hypothesis**.

It explained that viral conjunctivitis and dry eye syndrome can follow a respiratory infection and that both could explain my light sensitivity, excessive tearing, and gritty feeling.

It helped me evaluate treatment response, compare morning and evening photos, and assess whether I might benefit from restarting a previously paused antibiotic.

It didn’t make decisions for me, but it did something more powerful: **it made sense of my experience.**

It mentioned that artificial tear drops can help provide symptom relief. I took matter into my own hands and bought Systane Ultra on Friday, June 11th night itself.

When I applied **Systane Ultra**, an OTC artificial tear, and felt immediate relief, ChatGPT helped confirm that dry eye was likely contributing to my pain. When the pain returned the next morning, it calmly helped me assess whether the regression was due to environmental triggers (dry, hot air), withdrawal of antibiotics, or simply the natural course of viral recovery.

When I speculated on more complex hypotheses (e.g., increased intraocular pressure or referred pain from sinus pressure), ChatGPT didn’t dismiss me—it evaluated the plausibility of each scenario with clarity and clinical reasoning.

I continued using Systane and by Monday, June 16th, I had no aggravating symptoms, no discomfort, no pain except for a slightly red eye.

I visited an eye specialist on Tuesday, June 17th. She did all the tests and mentioned that this was most likely a follicular conjunctivitis.

However, she also prescribed Maxitrol eye drops which is a combination antibiotic and corticosteroid. I applied it for a day and discontinued it due to severe side effects (itching/ rashes/ blurry vision).

I stopped all meds and a couple of days later, all symptoms receded and I was on my path to getting better.

Looking back it was most likely a viral conjunctivitis that cleared up by itself in 2 weeks. But it was 2 weeks of pure hell.

**A New Model of Care?**

I want to be clear: ChatGPT didn’t replace my physician. But it filled a **critical vacuum in the healthcare journey** that many of us know all too well:

  • The confusing weekends.
  • The fear of escalating symptoms.
  • The lack of triage guidance.
  • The ambiguity between seeing a PCP, a specialist, or going to urgent care.
  • The days between appointments when symptoms shift and anxiety creeps in.

In these moments, ChatGPT became my **thinking partner, medical interpreter, and emotional buffer**.

It reminded me that dry eyes can appear suddenly after illness or due to medications. That OTC options are generally safe even if the root cause is viral. That deep, dull pain around the eye needs monitoring but isn’t always an emergency. It reassured me when things improved and gently encouraged medical follow-up when needed.

**My Takeaway**

Healthcare needs transformation not just in diagnostics or treatments, but in **how patients are supported between and beyond appointments**. We don’t just need faster doctors—we need better scaffolding for the whole journey.

For me, that scaffolding came in the form of an AI.

So yes—Dr. ChatGPT saw me. And sometimes, feeling seen is the first step toward feeling better.
#Local LLMs & GenAI#Healthcare & Life Sciences#Business Intelligence & Analytics