Why your 3 doctors aren't talking to each other

Today's Tip: The 7-Pill Problem

Your pharmacist has 90 seconds. Your doctor has 12 minutes. AI has all the time in the world to look at your pill bottles together.

Let me explain…

My friend Henry coached high school football in Pensacola for 34 years. He is 71 now, and like a lot of men his age, he sees three different doctors. A cardiologist for his blood pressure. A primary care doctor for everything else. And a sleep specialist who put him on something for restless legs last winter.

Each of those doctors prescribed something. None of them looked at the full list together. Henry was taking seven medications and two supplements, and nobody in the room had ever sat down and reviewed all nine at once.

He had been feeling foggy. Tired in the mornings. Off-balance once or twice when he stood up too fast. He chalked it up to age… and that is exactly what most people do.

One Sunday afternoon he sat at his desk with all his pill bottles in front of him. He typed the names and doses into Claude — every prescription, every over-the-counter, every supplement, including the magnesium his wife had bought him. He told the AI his age, his weight, and the symptoms he had been ignoring. Then he asked one careful question.

What came back was a polite, plainly-written summary. Two of his medications were known to amplify each other's sedating effect in older adults. His magnesium supplement was likely contributing to the morning grogginess. And the timing of one blood pressure pill, taken at night, was a known cause of the dizzy spells he had been having on the way to the bathroom.

He printed the summary, took it to his primary care doctor on Tuesday, and walked out with a revised schedule. Within ten days, the fog lifted.

Here is the prompt Henry used. It works for anyone taking more than two medications, and it takes about five minutes to fill in.

Notice what is happening here. You are not asking the AI to be your doctor. You are asking it to do the homework your doctor does not have time to do… and to hand you a list of careful questions to bring to your next appointment.

That distinction matters. Doctors love informed patients. They do not love patients who march in announcing that the internet told them to stop taking their statin. There is a huge difference between the two, and the prompt above is built to keep you on the right side of that line.

Now, I know what you are thinking. "I am not comfortable typing my medical information into a website. What about privacy?"

That is a fair concern, and you should take it seriously. You do not have to use your real name. You do not have to give your date of birth. The AI does not need to know who you are… it only needs to know your age, your weight, your conditions, and your medications. Use a free account, give a first name only or no name at all, and skip anything that feels too personal. The medical review still works.

And do not skip the most important step — printing the summary and bringing it to your doctor or pharmacist. The AI is the research assistant. Your doctor is still the one who decides.

Try this today. Get your bottles out. Fill in the prompt. Print the result. Then call your pharmacy and ask for a 15-minute medication review — most pharmacies offer them for free under Medicare, and you will arrive with the smartest questions they have heard all week.

It is the kind of thing nobody tells you to do. So we are telling you.

★ Quick Wins

Skill Builder: Getting AI to Cite Its Sources

There is one habit that separates people who use AI well from people who get burned by it. The people who use it well always — always — ask the AI where it got the information.

It is the simplest possible upgrade to how you ask questions. And it works because the AI knows the difference between something it is confident about (a well-documented drug interaction, a settled tax rule, a chapter from a textbook) and something it is guessing at (a brand-new medication, last week's news, your specific situation). When you ask it to cite, you force it to show its work.

Here is what a non-cited answer looks like… "Yes, that medication can cause dizziness." Useful, but you have no idea how confident the AI is. Here is what a cited answer looks like… "According to the FDA label and a 2023 review in the Journal of the American Geriatrics Society, this medication causes dizziness in roughly 8 percent of older adults, particularly when combined with [other drug]." Same answer. Twenty times more useful.

The trick is adding three words to the end of any question that matters. Save this prompt — it is a habit that will protect you for the rest of your AI-using life.

Try it once today. Take any question you would have asked the AI anyway and tack the prompt on at the bottom. You will be surprised how often the AI quietly admits that part of its answer is shakier than it sounds.

That admission is not a weakness. It is the most valuable thing the AI can give you. It tells you exactly where to do a little more checking… and where to relax and trust the answer.