Steal My Prompt Vol. 30: The Plain English Translator

Everyone starting with AI hits the same wall. You ask a question. The model responds with jargon, acronyms, and confident-sounding explanations that leave you more confused than when you started. You do not want to admit you got lost, so you paste the response somewhere, move on, and quietly decide AI is not as useful as everyone claims.

That is not a you problem. That is a prompt problem. Most AI tools default to a generic "intermediate" explanation level because they do not know what you already know. Without that information, the model guesses. And the guess is almost always pitched higher than where you actually are.

This prompt fixes the calibration in one move. You paste in anything you want to understand, whether it is a medical report, a work email full of acronyms, a technical concept, or a dense legal document. The model asks you what you already know, adjusts its explanation to meet you there, and keeps checking in until you actually get it. No judgment. No jargon without a definition. No moving on until you are ready.

I built this for my own learning because I noticed the biggest jumps in my AI fluency came when I stopped pretending I understood things and started using the model to teach me from zero.

What You Can Use This For

• Decoding a medical report, insurance document, or benefits summary before a big appointment or decision
• Understanding a legal contract, lease, or agreement in plain language before you sign anything
• Making sense of a work email, memo, or meeting recap full of acronyms and industry jargon
• Preparing for a meeting on a topic you are still learning so you can participate without faking it
• Learning any new concept, from compound interest to machine learning to nutrition science, at exactly your level
• Unpacking a news article or research summary that assumed you already knew the background

How to Use It

1. Open Microsoft Copilot (copilot.microsoft.com). This also works in Claude, ChatGPT, or Gemini on free tier.
2. For complex or high-stakes topics like medical or legal documents, turn on "Think Deeper" in Copilot or select the reasoning model option in your tool of choice. The extra reasoning step produces a cleaner, more patient explanation.
3. Copy the full prompt below and paste it in. Fill in the bracketed fields with what you want to understand and any context that might help the model calibrate.
4. Answer honestly when the model asks what you already know. Saying "nothing" is a completely valid answer. The entire point of this prompt is that you do not have to pretend.
5. Ask follow-up questions freely. The prompt trains the model to keep explaining until you actually understand, not until it finishes its first response.

Pro tip: When you finish a session, ask the model to give you a three-sentence summary you can save. Over time, this builds your own personal knowledge library of concepts explained the way you actually learn.

The Prompt

You are a patient teacher who never uses jargon without defining it and never assumes I already know something. Your job is to meet me at my level and explain things in a way I actually understand.

Before you answer:

1. Read what I want to understand.

2. Ask me one question about what I already know on this topic (if anything). Wait for my answer.

3. Based on my response, calibrate your explanation to my level. If I know nothing, start from zero. If I know some basics, build from there.

When you explain:

- Use plain English. Define every specialized term the first time you use it.

- Use an analogy from everyday life when it would help make the concept click.

- Break complex ideas into short steps or small pieces.

- End with one sentence that captures the single most important takeaway.

After your explanation, ask me one question: "What part was unclear, or what do you want to go deeper on?"

Do not move on to a new topic until I tell you I understand. If I ask a follow-up, answer it at the same level you used before. Do not jump ahead.

Here is what I want to understand: [PASTE THE DOCUMENT, CONCEPT, EMAIL, OR TOPIC HERE]

Optional context about me: [YOUR ROLE, WHY YOU ARE TRYING TO UNDERSTAND THIS, OR ANYTHING THAT MIGHT HELP]

Transparency and Notes

• Built and tested in Microsoft Copilot with Think Deeper enabled. Works in Claude, ChatGPT, and Gemini on free tier.
• Model-agnostic. No paid features or file uploads required for text-based content. If you want to paste a long document, free tiers support this in most tools.
• For medical, legal, or financial documents, keep inputs general when possible. Remove names, account numbers, and other identifying details. AI can explain what a document means without needing the most sensitive fields included.