Genius B*tchwork - Part 25: How AI Spotted the Pattern a Patient's Doctors Missed
A person fed years of symptoms and lab results into AI, and it connected the dots their doctors had missed. A real doctor then confirmed it. Here is how to use AI to bring better questions to your own doctor.
This is the Genius kind of B*tchwork post.
Most weeks I show you the boring work my own AI did for me. In these Genius posts I show you something smart that AI did for someone else. A real person, a real story, a real win. Then I hand you the exact steps so you can do a safe version of it yourself.
I want to be very clear before we go one inch further, because this one is about health.
This is not about getting a diagnosis from a robot. It is about getting better QUESTIONS to bring to a real doctor. AI is not your doctor. It can be wrong. It cannot examine you. Nothing here replaces a real visit with a real human who went to medical school. Read that line again. I am going to repeat it a few times on purpose, because it matters more than anything else in this post.
Okay. Here is what happened.
A person had been sick for years. They had a pile of symptoms that did not seem to connect, and a stack of lab results from different visits. Their doctors had looked at things one at a time and had not landed on an answer. So this person took years of their own symptoms and their own lab numbers and pasted them into an AI. They asked one simple thing: what could explain all of this together?
The AI looked at everything at once. It noticed a pattern across the symptoms and the labs that pointed at one possible cause (a gene issue called MTHFR). It was a thread their doctors had not pulled on. The person took that idea to a real doctor, who looked into it and confirmed it.
That is the part that matters. The AI did not diagnose them. It gave them a better question. A real doctor did the actual confirming. (Link to the story at the bottom.)
And here is the thing nobody tells you. Doctors are rushed. They often see your problems one visit at a time, not all together. They almost never have your whole history laid out in front of them in one view. So a pattern that only shows up when you line everything up can get missed. Not because anyone is bad at their job. Because the system gives them ten minutes and a tired brain.
You can do the lining-up part yourself. With AI helping. To walk in with a sharper question.
What doing this the old way costs you
If you have ever been the sick person nobody can figure out, you know the loop.
You go to a visit. You try to remember every symptom while a busy person looks at the clock. You mention three things, forget two, and the doctor focuses on one. You leave with a guess, or a shrug, or "let's wait and see." Next visit is months later with a different doctor who does not have the full picture. Nothing ever gets connected because nobody is looking at all of it at once.
That is the trap. Your story is spread across years and across people. So the pattern never shows up.
AI is good at exactly that one job. It can hold your whole messy history in view at the same time and say, "hey, these things tend to go together, ask about this." It does not diagnose you. It hands you a smarter question to bring to a human.
What you get in this one
I broke the whole thing down so a total beginner can copy it safely. By the end of this post you will have:
The exact way to gather your symptoms and labs so AI can actually use them.
The copy-paste prompt that makes AI look for patterns across everything.
The prompt that turns those patterns into questions for your doctor.
The steps in order, plus the hard safety rules so you do not hurt yourself with this.
You do not need to know anything about medicine. That is the whole point. You are not trying to be the doctor. You are trying to walk into the doctor's office with a better question than "I just feel off."
Quick and honest note before we start
I am not a doctor, and your AI is for sure not a doctor. This is a way to get better QUESTIONS to ask a real one, nothing more. AI can be confident and wrong. It can scare you over nothing, or miss something real. Never start, stop, or change any medicine, test, or treatment based on what an AI says. Always confirm everything with a real doctor before you act. And take your personal info off anything before you paste it. We will cover that.
Behind the paywall, you get the full how-to:
The 5 steps in order, start to finish.
Both copy-paste prompts, ready to use.
Exactly what to gather, and what to leave out.
How to turn the AI's output into a short list of questions for your visit.
The do-not-do list, so you stay safe and do not panic over a robot's guess.
Subscribing unlocks the rest of this exact post right now, not just future ones. If you or someone you love has been the person nobody can figure out, this is the post that helps you walk in with a real question.
The full walkthrough is below the paywall.
Click the button, subscribe, and the locked section below opens right away. You get this article now.
