Session 6.8: The "Prompt Me Things" Process

Course → Module 6: Voice Capture & Preservation

Session 8 of 9

Inverting the Flow

The standard AI content workflow: AI generates text, you edit it. The "Prompt Me Things" process inverts this. You provide the raw substance. The AI assembles it into polished prose. Instead of the AI guessing what you know, you tell it explicitly. Instead of editing AI-generated ideas, you generate the ideas and let AI handle the sentence-level assembly.

This inversion solves the deepest problem in AI content production: authenticity. AI can generate text about any topic. It cannot generate text that contains your specific experience, your particular insight, your hard-won lessons. But you can provide those things as raw material, and AI can weave them into readable prose while preserving your voice.

The "Prompt Me Things" process turns you into the source, not the editor. AI is excellent at organizing ideas into paragraphs, creating transitions, and maintaining consistent structure. AI is terrible at having ideas worth organizing. You provide the ideas. AI provides the prose. Each party does what it does best.

The Process

graph TD A["Step 1: AI generates
targeted questions
about your topic"] --> B["Step 2: You answer
each question honestly,
in your own words"] B --> C["Step 3: AI organizes
your answers into
a coherent structure"] C --> D["Step 4: AI writes
polished prose using
your raw answers
as source material"] D --> E["Step 5: You review
for accuracy and
voice consistency"] style A fill:#222221,stroke:#c8a882,color:#ede9e3 style B fill:#222221,stroke:#6b8f71,color:#ede9e3 style D fill:#222221,stroke:#c47a5a,color:#ede9e3 style E fill:#222221,stroke:#6b8f71,color:#ede9e3

Designing Effective Questions

The questions must extract specific, experiential knowledge. Generic questions produce generic answers. The right questions force you to draw on experience that only you have.

Bad Question (Generic)	Good Question (Experiential)
"What is project management?"	"What is the worst project management advice you ever followed, and what happened?"
"Why is quality important?"	"Tell me about a time quality control caught something that would have been embarrassing."
"What should beginners know?"	"What did you get wrong for the first two years that nobody told you about?"
"How do you handle clients?"	"Describe a client interaction where you said no, and what happened to the relationship afterward."
"What trends do you see?"	"What is one thing everyone in your industry believes that your experience contradicts?"

How to Answer

Answer in whatever format is fastest for you. Voice memos that you transcribe. Bullet points in a text file. Stream-of-consciousness typing. The answers do not need to be polished. They need to be honest and specific.

Good answers contain: specific details (dates, numbers, names), emotional honesty ("I was terrified"), contrarian opinions ("Everyone says X. In my experience, the opposite is true"), and concrete examples rather than abstract principles.

Bad answers are: abstract ("Quality is important because..."), hedged ("I think maybe sometimes..."), or generic ("Most people find that..."). If your answer could appear in a textbook, it is not specific enough.

From Raw Answers to Polished Content

Feed your raw answers to the AI with your voice documents (persona.md, design.md, style.md) as context. The prompt: "Using the raw answers below as source material, write a [content type] that preserves the specific details, experiences, and opinions. Do not add information not present in the raw answers. Do not generalize specific examples."

The constraint "do not add information not present in the raw answers" is critical. It prevents the AI from padding your specific experience with generic filler. Everything in the output should trace back to something you said.

When to Use This Process

The "Prompt Me Things" process works best for content that depends on personal experience: opinion pieces, case studies, lessons learned, industry commentary, and any writing where your authority comes from what you have done rather than what you have read. For purely informational content (tutorials, how-to guides), the standard pipeline from Module 8 is more efficient.

Assignment

Create a list of 20 questions about your area of expertise that only you can answer. Not generic questions, but experiential ones that force specific, personal responses. Answer 5 of them in your own words (raw, unpolished, in whatever format is fastest). Feed those answers to AI with your voice documents as context and the instruction to write based only on your raw material. Evaluate the output: does it contain your specific experience? Does it sound like you?

The "Prompt Me Things" Process