A creative non‑fiction retelling, true in every detail.
I’ve been focused on fiction lately, but something happened this past month that nudged me into writing this. Call it creative non‑fiction, call it an exception to my rule. Either way, it’s a story I didn’t expect to tell, let alone witness, especially in a time when people seem eager to throw stones at both AI and Boomers.
It began with a health issue that was driving both of us a little crazy, though my wife bore the brunt of it.
Every afternoon, her lower legs, the shins and feet, would become strangely sensitive to cool air.
Not cold. Just an uncomfortable, almost painful tingling. And her skin wasn’t cold to the touch. Even the lowest setting on the ceiling fan could keep her awake at night.
We did what most people do: we guessed.
Poor circulation?
Some kind of nerve issue?
Maybe just one of the many consequences of getting older?
She tried warmer socks, thicker blankets, adjusting the air conditioner in endless combinations. Nothing explained it. Nothing helped.
Then she started talking to ChatGPT.
I didn’t think much of it. Like a lot of people, I assumed AI was just a glorified search engine with better manners.
I couldn’t have been more wrong.
What impressed me wasn’t that it gave her an answer, because it didn’t. Instead, it asked her to pay attention.
To observe. To write things down. To look for patterns.
Every day she’d report what happened: the time the symptoms appeared, whether she’d been walking, the room temperature, the humidity, whether the fan or AC was on, how long she’d been sitting. Sometimes she’d return an hour later and say, “I think I described that wrong,” and correct herself.
The AI adjusted.
The next day, they picked up where they left off.
I watched this with my own curiosity. It didn’t feel like googling something. It felt like watching two people quietly run an experiment together.
There was one moment that changed how I looked at AI. Eventually, ChatGPT suggested she see a doctor, not because it had the diagnosis, but because it could see where the uncertainty still was. It even helped her figure out what questions needed answering and what kinds of tests might help.
Before the appointment, she asked something unexpectedly sweet:
“Would the doctor be offended if I showed him my notes that AI helped organize?”
The answer was simple: doctors care about clear medical history, not who helped assemble it.
She showed him.
He took the time to read everything carefully, did the exam, ordered the tests.
When the results came back, almost every major area ChatGPT had flagged as needing clarification had been checked by the cardiologist. Not because AI replaced him, but because good medicine, like all good science, follows the evidence.
Oddly enough, that experience didn’t make us trust AI more.
It made us appreciate the value of a doctor.
For the first time, we understood why those tests were being ordered.
Something else surprised me. I have heart disease myself, two coronary stents, and for years I believed lower blood pressure was always better. That’s what we’d been told by our nurse practitioner back in Canada.
So when our cardiologist in Thailand adjusted her medication, introduced a couple of focused supplements, and her numbers rose a little, I got worried. Wouldn’t you?
ChatGPT didn’t tell us the doctor was right. It simply explained how guidelines have changed. The goal isn’t to push blood pressure as low as possible, near fainting, again, as we’d been told. It’s to help the patient actually live well.
Over the weeks, her blood pressure stabilized.
She started sleeping normally again.
Our morning walks returned.
Then our swimming routine resumed.
Her symptoms slowly faded.
Online, people ask me, “What’s the secret to getting good results from AI?”
My answer is always the same: repeated iteration.
Don’t expect one perfect prompt or one perfect answer.
Work with it.
Observe.
Correct it.
Ask again.
Think again.
Repeat.
The quality comes from the conversation, not the first question.
Looking back, the most amazing part wasn’t that her symptoms improved, though that was a huge relief. The remarkable part was watching a 59‑year‑old woman become genuinely curious about an entirely new way of thinking.
She never handed over responsibility for her health.
She still saw her doctor.
She still trusted sensible medical evidence.
She still made her own decisions.
AI didn’t replace any of that.
It simply became another thoughtful, if unique, partner in the process.
I sometimes joke that she spent a whole month giving ChatGPT a daily health report. She laughs. But together they built something neither of them had on day one: understanding.
Not certainty.
Not magic.
But understanding.
And maybe that’s the most valuable thing AI has to offer.
If I learned anything from watching all this, it wasn’t about artificial intelligence or any particular model. It was about curiosity.
My wife has always been curious.
AI didn’t give her that.
It simply gave her a new companion to explore the unknown with.
Watching that partnership grow has been one of the most unexpectedly hopeful things I’ve seen all year.
Note: This is a true account of how we used AI as a thinking partner and not a diagnostic tool. Any medical decisions should be made with a licensed healthcare professional.