2026 September 01 AI as your health assistant

Hi, this is Jim Cranston from 7EveryMinute and 7EveryMinute.com, the podcast and website about reimagining your life. Thanks for joining me to talk about real-life AI as we age. If you like what you hear today, please leave a like, subscribe, tell your friends, send me a message.

This week, we're going to be talking about using one of the major AI models for real medical diagnosis assistance. As a preamble, I'm not talking about replacing your care team or self-diagnosing. I'll be talking about a way to organize your medical history so that you can easily keep track of it, be on the lookout for possible interactions so that you can speak to your care provider about it, and allow you to look at your health in ways that normal Western medicine tends not to do, such as weaving in things like meditation, specific exercises, or other care practices.

So I'm going to start with example one tonight, and this is an area where AI really tends to shine, similar to some other work I did where I was summarizing some things — I talked about that a little bit last week. Tonight we're going to start with my dog, Rocket, and the patient — we'll describe the patient.

He's a seven-year-old mixed breed rescue dog that had an onset of pain while defecating. The obvious culprit: intestinal foreign body. Unfortunately, that's not the answer.

So what has followed is working with a number of very capable veterinarian specialists, all of whom have been able to thoroughly rule out common and even less common causes. But this has taken months, and the patient's still in pain, so it's ideal to continue trials, right? We have to make some progress. Tonight I spent about an hour working with Claude Opus, and since it's important to work with a subjective outcome, I'm going to repeat this process with ChatGPT probably later tonight or tomorrow morning.

I was able to give Claude Opus every symptom, action, observation, test result, test performed, medications prescribed, and the resulting outcomes of those medications, the timeframe after the medications, the patient's origin story, our current location, the seasonal onset — every single thing you can think of, I could pop into AI. And I can't make this clear enough: I'm totally satisfied with what my specialist has been able to accomplish. They did wonderful things.

But the patient is only one of many of their patients, and I have the luxury of spending a couple hours just pumping data into each of the AI systems and answering the little detailed questions they ask me back. The results thus far are pretty promising. Plus, the AI is going out of its way saying, "This is not a diagnosis, these are potential areas of exploration to be shared with the care team," and went further to offer to generate a summary document with the reasons for the suggestions. So am I saying this is solved? Not yet.

But as I hoped, all the little hints taken together over a four-month period actually do resolve to some potentially actionable items. And to be clear, the tests that were done absolutely ruled out a number of likely and more commonplace diagnoses. So it wasn't a waste of time or money.

Now think about your own health and your own care team. Chances are they're very busy, especially if your first stop is a family medical practitioner or GP. So how do you coordinate your care across various caregivers and specialists, the medications they all prescribe, the reasons for the medications — let alone all the lifestyle questions, such as: What does your diet look like? How active are you? How would you rate your level of stress and contentment? How many social connections do you have? Have you traveled recently? To where? How did you feel? Did you get enough sun? Is your room dark when you try to sleep? Do you turn off devices an hour before bed? How long before sleeping do you eat? And the thousands of other things that have a profound influence on your health.

Think about all the stuff I just mentioned — that doesn't even scratch the tip of the iceberg. So one option is to use AI to really track all of your medical information. Let me stop right here and say: each AI system has its own privacy rules, and free plans generally have little to no privacy.

So read the privacy policies, ask the AI to explain them to you, ask a different AI to explain what the other AI's privacy policies are, and see if you're comfortable with the answers. Tonight I'm going to be talking about using Claude Opus, one of the models from Anthropic, as opposed to OpenAI, which has ChatGPT — and there are other models as well. With Claude Opus, you may have to gather some things together yourself, but this gives you a bit more flexibility — with power comes responsibility, so you have to do a little more work. Alternatively, you can use OpenAI's Health in ChatGPT option. This automates some of the steps, though it isn't quite as flexible — I'll talk a little more about that toward the end of the podcast.

So no matter which you use, if you're comfortable, start by finding all your old medical test reports. It's such a mess — no interoperability. I've started seeing startups that offer a single interface bringing all your medical information together, which is a sad state of affairs in itself. But nonetheless, you can usually log into a portal for each of your doctors or labs and get the results yourself. If you can't log in, torment their help desk until you can get your data. You have the right to see it, so gather it all together. Get all your labs for as long as you can, because trends count more than a single report.

Then take any diagnoses and care plans — put all of those things, the labs, the diagnoses, the care plans, prescriptions, all of that, into files, or a single file, in a single directory, and ask the AI to generate a skill — that's why it's called a skill — named "My Medical History." Then ask the AI if there's any other medical information it would like from you, and have it interview you and update that My Medical History skill file.

Now put in everything related to lifestyle that comes to mind, such as what activities you do, if you exercise and how much, how active you are, how you eat, what you eat, do you take any supplements and which ones, and all the social things you do, what groups or organizations you belong to — just everything you can think of. Again, ask the AI to generate a second skill, this time named "My Lifestyle History." As before, ask the AI to interview you about any missing lifestyle information — it might even want pictures of you at different ages — and have it update the My Lifestyle History skill file.

And voilà, you now have the basis for working with one or more AI systems, and they will start with the full knowledge of your medical and lifestyle history. You can then ask them specific questions without having to start each time with a bunch of typing, or forgetting to put something in. Also, every time something changes — like a new test comes in — you can ask that same AI to update the relevant medical or lifestyle history file with the new information. It knows what the file is, it'll open it, update it, and tell you that a new version is available.

Think about what you just created. It's a functional definition of how you live your life that you can share with the AI — and with your real doctors, if you want — so that when something seems unusual, you can simply ask them about it. Very important, again: this isn't to replace a medical provider, only to look across all your care providers plus your own lifestyle, so that you're better able to give good information to the real human doctors. It also allows you to consider things and ask questions about health-related options besides just medicines, such as activities and wellness options, meditation, unconventional exercise options like dancing, and interesting activities in your area that match your needs, et cetera.

One of the huge problems with the current state of Western medicine is that it's turned into a business, and businesses have to have a focus, of course. And that business isn't the business of keeping you well — it's the business of getting you better after you get ill. There's no secret conspiracy here, but those are two separate things, insofar as Western healthcare systems have developed. Yes, I know there are HMOs and they were supposed to address that, but honestly, most of those are so busy just addressing basic care issues that they have little time to address the real lifestyle issues. This is an area where AI can make a huge difference.

Imagine a planning partner that not only knows your personal preferences, but also all the things that may be affected by both your health and lifestyle realities. Imagine there's a great event coming up, but it's at a seafood restaurant, and you have severe seafood allergies — it can warn you of that and suggest other options for that night, or techniques to mitigate the risk while you're at the restaurant. Or you just notice you can't balance the way you used to, and you want some exercises to help improve your balance. Well, the AI can help you with the exercises, as well as note some of the other things to look for to make sure it probably wasn't caused by a TIA, a stroke, and maybe suggest what you can ask your medical provider in case they want to see you.

This isn't radical, crazy talk. OpenAI's ChatGPT recently rolled out Health in ChatGPT on all plans, specifically with medical record integration, wearable and app syncing, and personalized guidance.

The reality is that medical knowledge continues to expand, and breakthroughs in healthcare — both medical and non-medical — are happening every day. It's not reasonable to expect that one person who sees you once a year, maybe when you get sick, is able to speak to a complete personalized health plan for each patient on each visit, with updates you never even told them about. But by taking control of your own health history, plus weaving in your own lifestyle history and desires, you're able to create that first step toward a life and health planner who is always available to bounce questions off of.

Please let me know if you're interested in doing a walkthrough or setting one of these up. Again, take a look at OpenAI's Health in ChatGPT option, or we can set it up manually with your own data in Claude Opus or another AI.

So, something to think about — that's it for the evening. I think it's actually a really interesting area, and I would encourage you to do some poking around on the internet. Ask about Health in ChatGPT, ask about how you can manage your health with any of the AI systems. Again, this isn't trying to replace your doctor, but it's about getting your information together so that one person can follow it — and that gives you a way to bring things forward to your doctor. I think you'd be surprised at how much difference this could make in your health, because I think one of the problems we have with so much of our healthcare is that all the necessary information isn't available to all the people involved when they need it.

So, enough on that. Don't forget, we have a freebie available — it's called Our Five Pillars of Life and Aging. It's a short video where I explain how our lives are shaped by five major influences and how we can balance and adjust those as we age. Check it out at 7EveryMinute.com/five-pillars-freebie. I think it's pretty interesting, and it plays into a lot of different things I'm working on right now — I think it has a lot of information you'll find useful.

Also, remember, as always, one of the best ways to care for yourself is to care for others. See UKR7.com — links to places that help the people of Ukraine, and WCK.org, World Central Kitchen, which works in disaster areas around the world — both do great things, slightly different focus. World Central Kitchen shows up anywhere there's disaster; they do amazing things. Those are both international sites, but there are also a lot of local charities always looking for help. And if you're not in a position to give right now — even just a simple smile can change someone else's day for the better in ways you can't imagine.

As always, thanks for stopping by. If you found something interesting or useful, please pass it along. Please subscribe, hit that like button, and leave a comment. Have a great week.

Remember to live the life that you dreamed of, because that's the path to true contentment.

Love and encouragement to everyone. See you next week on 7EveryMinute and 7EveryMinute.com. Thank you.