-
Four Ingredients of a Strong Medicare Prompt
Age or birth month, state or ZIP code, current coverage, employment status, and income year if cost is involved. Medicare rules bend on all of these.
Say what you are actually deciding, not just the topic. “Should I delay Part B while working?” beats “Tell me about Part B.”
Ask for a timeline, a decision table, a checklist, or a plain-language summary. Naming the format changes the quality of the answer.
Ask for official sources, ask it to flag state-specific variation, and ask it to list its assumptions and what it could not verify.
-
Weak Prompt vs. Strong Prompt
“Explain Medicare Part B to me.”
Returns an encyclopedia entry you could have found anywhere.
“I turn 65 in March and still work for a 40-person employer with an HSA plan. Give me a month-by-month timeline of what I can and cannot do with Part A and Part B, the penalty risk of each choice, and cite the Medicare.gov pages.”
Returns a dated plan you can act on and verify.
-
Why Prompting Matters
A vague question gets a generic answer. Medicare is never generic — your ZIP code, your birthday, and your prescriptions change everything.
Why Context Matters
Tell the AI what you have, where you live, and what you are deciding. Those three facts turn a Wikipedia summary into an actual answer.
The Three-Fact Rule
Always ask for the source. If a Medicare answer can't be traced back to Medicare.gov, SSA.gov, or your plan documents, treat it as a starting point, not a decision.
Verify Before You Act
Ask the AI to show its assumptions. The assumptions are where the errors hide.
Surfacing Assumptions
-
Common Pitfalls
