Tell both the same facts, but frame them for what each one can do. A clinician can examine you, order tests, and make a clinical judgment, so lead with your main concern and a tight timeline, then let them steer. An AI assistant cannot examine you or take responsibility for your care, so give it fuller context, ask it to help you think and prepare rather than to conclude, and carry anything important back to a clinician.
This page is for anyone who will tell the same lived facts to both: before a visit, after one, or between visits. The mistake in each direction is real. Treating an AI like a doctor invites false confidence. Treating a clinician like a chatbot wastes a short appointment on context they did not need.
Start from one fact list
Do not build two separate stories. Build one list of facts, then frame it twice. The list is the same either way: what you noticed and when, what changed around that time, medicines and diagnoses already known, sleep and load, what you already tried, and your questions. If you do not have that list yet, Bring-this sheet: a pre-visit / pre-AI clue checklist gives you a template. If your notes tend to drift into labels, Observe and record without diagnosing or coaching helps clean them first.
If Alzheimer’s and ADHD are part of your situation, start from the page that fits. Caregivers running a household day for someone else can use When Alzheimer’s care and ADHD share one household day. People who have both conditions themselves can use When Alzheimer’s and ADHD live in the same body: prep before you speak.
Keeping one fact list has a practical benefit. If the clinician and the AI later seem to disagree, you can check whether they heard the same facts. Often they did not.
What changes in the clinician frame
Time is short, and judgment is theirs. A clinician brings examination, tests, history, and accountability. Your job is to give them the clearest signal fast and to make sure your top question gets answered.
- Lead with the concern, not the backstory. “The main thing: I’ve been waking at 3 a.m. most nights since early September.” Then the timeline.
- Give the timeline compressed. Three to five dated lines. Offer the full sheet if they want it.
- State facts already known exactly. Medicines as on the label, with dose and timing. Diagnoses, and who gave them.
- Say what you have tried and what followed. Briefly, with dates.
- Ask your ranked questions. If time runs short, ask question one first.
- Do not lead with an AI conclusion. “An AI said it might be X” can anchor the conversation. If you want to raise it, phrase it as a question: “Is X worth considering?”
- Write down what they said. Next steps, what to watch for, when to come back.
What changes in the AI frame
Time is long, and judgment is not its job. An AI assistant can read everything you give it, take its time, and help you organize, phrase, and prepare. It cannot examine you, it may be wrong in confident language, and it is not responsible for your care. That shapes how you talk to it.
- Give fuller context. The AI does not know you. Include age range, relevant history, the full timeline, and what you already tried. Leave out identifying details it does not need, such as full names, addresses, and ID numbers.
- Say what you want from it. “Help me prepare questions for my appointment” or “Help me organize these notes by date” works better than “What do I have?”
- Ask for possibilities and questions, not a verdict. “What kinds of things might a clinician consider, and what would they likely ask me?” keeps the AI in a supporting role.
- Ask it to flag urgency. “Is anything here the kind of thing that should not wait for a routine visit?” If the answer is yes, or if you are worried, contact a clinician or emergency services. Do not keep chatting.
- Do not ask it to adjust medicines or treatment. Dose, timing, and starting or stopping anything are for the prescriber.
- Ask what it is unsure about. Then mark those items “verify with clinician.”
Some agents can bring in outside perspective. For example, Life Node is a module being built for the agent someone already uses. It indexes other people’s shared, anonymous experience as perspective, and it does not advise or diagnose (background). Perspective like that belongs in the AI frame as context. It is not a substitute for the clinician frame.
Takeaway: one fact list, two frames
Fill in the left side once. Use the middle and right columns to decide how each fact is presented.
| Fact (from your sheet) | Clinician frame | AI frame |
|---|---|---|
| Main concern | Say it first, in one sentence. | State it, plus what you want help with (prepare, organize, phrase). |
| Timeline of observations | 3–5 dated lines; offer the full sheet. | Full dated timeline, oldest first. |
| What changed around then | Mention only if it seems relevant. | Include it all; ask what a clinician might connect. |
| Medicines and known diagnoses | Exact names, doses, timing, who diagnosed. | Same facts; say “do not suggest changes.” |
| Sleep and load | One-line summary; device data if asked. | Fuller numbers with device and dates. |
| Already tried | What, when, what followed — brief. | Same, with more detail. |
| Questions | Max 3, ranked; ask #1 first. | Ask it to help sharpen your questions for the visit. |
| Urgency | Ask directly: “Should this wait?” | Ask it to flag anything that should not wait; then contact a clinician. |
| After | Write down next steps they set. | List what to verify with a clinician. |
Copy-paste openers:
CLINICIAN
"The main thing I want to understand today is ____.
It started around ____. Here are the key dates: ____.
I'm currently taking ____ as prescribed. I've tried ____ since ____.
My first question is ____."
AI ASSISTANT
"I'm preparing for a clinician visit, not asking for a diagnosis.
Context: ____. Timeline (oldest first): ____.
Medicines as prescribed (please don't suggest changes): ____.
Please help me (a) organize this, (b) sharpen up to 3 questions,
and (c) flag anything that sounds like it should not wait."
This page is general information about communicating health facts. It is not medical advice. An AI assistant is not a clinician; diagnosis and treatment belong with qualified professionals.