Module 06 · Tier 2 — Basic Prompting

AI for Writing & Communication

Make AI your thinking partner, not just a text generator.

12 min Tier 2 — Basic Prompting

Why This Matters

Writing with AI isn't about replacing your voice — it's about removing the friction between your ideas and the page. Most people use AI as a typist: tell it what to write, paste the result. The better use is as a thinking partner: someone to push back on your ideas, help you find what you're actually trying to say, and give you a first draft that you can make your own. This lesson is about the difference.

The Concept

The Drafting Trap

The most common writing mistake with AI: asking it to write something fully formed before you've thought through what you actually want to say. AI is excellent at producing text that looks finished. That polished surface hides a problem: you haven't actually done the thinking.

The output sounds authoritative. It's well-structured. It might even be accurate. But it's not your thought — it's AI's best guess at what someone in your position might write. The thinking that should have happened in the drafting process never happened.

The better pattern is to use AI in the thinking phase, not just the writing phase.

AI as Thinking Partner

Some of the highest-value writing uses of AI don't produce a single word of your final document:

"What am I actually trying to say here?" — Paste your rough notes or half-formed thoughts and ask AI to identify what the core argument or point seems to be. Then decide if that's right.
"What are the strongest objections to this?" — Ask AI to steelman the opposition to your argument. The pushback often clarifies your position.
"What's missing?" — Share your outline or draft and ask what questions a skeptical reader might have that you haven't answered.
"What order makes sense?" — Dump your bullet points and ask AI to suggest three different ways to sequence the argument. Seeing different structures often reveals the right one.

The Voice Problem (and How to Solve It)

AI-generated text sounds like AI-generated text. The vocabulary is slightly too formal. The transitions are too smooth. The qualifications are too balanced. It reads like a competent committee wrote it rather than a specific person.

How to get AI to write in your voice instead:

Give it samples. Paste two or three examples of your own writing and say: "Match the tone and style of these examples."
Give it adjectives. "Write this in a voice that is direct, conversational, and slightly irreverent. Not formal, not cautious."
Edit aggressively. AI's first draft is your raw material, not your output. The edit is where your voice enters.

High-Value Writing Tasks by Category

Emails and messages: Draft → ask AI to make it clearer and more direct → edit back to your voice. Particularly valuable for difficult conversations — AI removes the emotional charge that makes hard messages harder to write.

Long-form writing: AI for outline and structure. You for the substance. AI for language cleanup after you've said what you mean.

Summaries: Strong AI use case. Paste the original, ask for a summary calibrated to your audience and purpose. Still verify key facts.

First responses in new situations: When you're writing something you've never written before — a new type of email, a genre you're not confident in — AI's draft gives you a template to react to, which is faster than starting cold.

The thinking partner in practice

Here's an example of using AI for the thinking phase of a document, not just the writing phase:

Prompt: "I need to write an email asking my team to adopt a new workflow tool that nobody asked for and that will add work in the short term before it saves time. Here are my rough notes: [paste rough notes]. What is the core argument I'm actually making? What are the three most likely objections? And what am I missing that would make this more persuasive?"

This prompt doesn't produce an email — it produces the thinking you need to write a good email. The email you write after this conversation will be far stronger than anything AI would have drafted without it.

Hands-On Exercise

Use AI for the thinking, not just the writing

ClaudeChatGPTGemini

Choose something you need to write this week — an email, a document, a presentation, a message. Before asking AI to write anything, ask it to help you think: 1. Share your rough notes or initial thoughts and ask: "What is the core point I seem to be making? What questions does a reader still have?" 2. Ask: "What are the strongest objections to this approach?" 3. Ask: "What's the most important thing I should say first?" Then write your own draft based on those answers. You can use AI for editing passes afterward — but write the substance yourself. Compare the result to how you'd normally write something with AI. Which approach produced writing you're more confident about?

This is a longer exercise. The point is to notice what happens to the quality of your thinking when you don't outsource it.

Active Recall

Before moving on — close this lesson and answer these from memory. Then come back and check. Testing yourself (not re-reading) is how this sticks.

1 What is "the drafting trap" in AI-assisted writing, and why does it matter?

2 Name three ways to use AI in the thinking phase of writing, before generating any draft text.

Reflection

Which types of writing in your work could benefit most from AI as a thinking partner rather than a text generator? What would you need to change about how you currently use AI for writing?

Key Takeaway

The highest-value AI writing use isn't drafting — it's thinking. Use AI to stress-test your argument, identify gaps, and find the right structure before you write. Edit aggressively to restore your voice. The thinking is yours; AI handles the friction.