Like many households, we have a running daily question that never quite gets easier to answer: what’s for dinner? It sounds trivial but it isn’t. Even though we typically plan out meals for the week and cook to have leftovers, trying to satisfy food preferences, nutrition goals, while also creating variety, can be exhausting.

I’ve been using a food tracking app called Cronometer, which I use to log my micro and macronutrients and calorie intake. I have it synced to Apple Health, so it also tracks my activity and energy expenditure too. One evening I had an idea. What if instead of asking Claude a generic question about meal planning, I gave it my actual data? My food history, preferences, my lab results, my goals. What happened next genuinely changed how I think about what AI is capable of.

Most people use AI the way they used Google: type a question, get a generic answer, move on. That’s useful, but it’s also a fraction of what’s actually possible.

Here’s what I did. I uploaded my historical Cronometer data into Claude, including months of food logs, micro and macronutrient patterns, calorie intake, and energy expenditure from my connected Apple Health app. Then I went a step further. I uploaded my demographic data, my last year of lab results, vital signs, and my current medications and supplements. I gave Claude the full picture of where I actually am, so it can use that data to get me where I want to be.

Then I told it our family’s preferences. We prefer chicken over beef, salmon is fine but no more than one to two days per week, not a fan of hummus, and a few other specifics. Claude produced several tailored meal options. And then it did something I hadn’t asked for.

It offered to create a Costco-friendly shopping list from the recipes it suggested. Not because I mentioned Costco, but because it had noticed the Kirkland brand on items in my uploaded food logs and made the inference on its own. I hadn’t prompted it. It just paid attention.

What happened next genuinely changed how I think. Claude flagged that one of my lab values was slightly outside the normal range. It was careful to tell me that it is not a doctor and does not make medical recommendations. But it suggested I might want to ask my doctor about a zinc supplement, because research supports its relevance to that particular result. It even went so far as to tell me that, if I do end up taking the zinc from the doctor’s recommendation, to time it where I am not taking it with another medication I am taking because they have similar absorption pathways and can interfere with each other’s uptake. That observation came entirely from Claude cross-referencing my lab data with nutritional research. I had not asked about it. It noticed on its own.

I then asked Claude to generate a document with specific questions to bring to my next doctor’s appointment, based on everything it had seen. I saved that document in my project files. After my checkup, I’ll feed the new results back to Claude so it can start recognizing trends over time and refine its recommendations for diet, exercise, and lifestyle accordingly.

What started as a solution to the dinner question has become an ongoing personal health system. One that updates as my data updates. One that connects dots I wouldn’t have connected, and ones that I wouldn’t even expect my doctor to connect from an annual checkup.

That is a fundamentally different use of AI than most people are even aware is possible.

The shift isn’t about which AI tool you use. It’s about what you give it. Generic prompts produce generic answers. Real context, with your actual data, situation, and specific goals, will produce something else entirely. The same principle applies far beyond meal planning. It also applies to your financial data, career history, learning patterns, personal goals. Most people are using AI at a fraction of its potential because they are still treating it like a search engine. The upgrade isn’t a new tool. It’s a new posture.

Three questions worth sitting with this week:

1. What data about your own life — health, finances, habits, goals — already exists somewhere that you’ve never thought to give an AI full access to?

2. Are you using AI to ask questions, or to build something that gets smarter about you over time?

3. What would change if your AI assistant actually knew your full picture, not just the question you typed today?

Three things worth your time this week:

📚  Read

Good Energy by Casey Means, MD — a rigorous and practical guide to metabolic health that puts lab results, biomarkers, and personal health data directly in the reader’s hands. Means argues that most chronic disease is rooted in metabolic dysfunction, and that the tools to understand and address it — including how to interpret your own bloodwork — are more accessible than most people realize. The doctor’s appointment document Claude helped prepare is the exact kind of informed, data-driven engagement Means advocates for. If Issue #7 changed how you think about using AI for your health, this book will change how you think about your health data.

🧠  Concept worth knowing

Context window — the amount of information an AI model can hold and reference at one time. The larger the context window, the more data you can give it in a single conversation. What made the Cronometer + lab results + medications upload possible is that modern AI models have large enough context windows to hold and cross-reference all of it simultaneously. This is a relatively new capability. A year ago, most of what this issue describes wasn’t practically possible for a consumer. It is now.

📊  Stat worth sharing

According to a 2024 survey by the American Medical Association, fewer than 1 in 5 patients review their own lab results before a doctor’s appointment — and most who do have no framework for interpreting what they’re looking at. Giving an AI your actual lab data and asking it to help you prepare questions isn’t replacing your doctor. It’s arriving as a better patient.

Zoom out. See what’s possible.

— Chuck

If this issue resonated with you, the best way to support The Wider Lens is simple: share it with one person who would find it valuable. Forward this email, send them the link, or share it on social. It means more than you know.

SOURCES

Cronometer food tracking app: cronometer.com
Apple Health: apple.com/health
Good Energy by Casey Means, MD: amazon.com/Good-Energy-Surprising-Connection-Metabolism/dp/0593712641