The One Prompting Technique Everyone Can Use

And why those 100-prompt guides are wasting your time

Teaching Kimmy/Claude about “The 5 Questions” (IYKYK)

Why it matters

After months of experimenting with AI writing tools, I've discovered most of us are making this way too complicated. The best technique is also the simplest.

The big idea

Add this to every important AI prompt:

"Before you respond, ask me 5 questions that will help you give me a better answer."

Why this works / The Science Behind It

We’ve entered the world of “Generative Thinking.” 

A terrific article by Michael Spencer and Nick Potkalitsky in the Substack “AI Supremacy” shows how this approach develops "prompt literacy" - helping writers craft queries that open possibilities rather than narrow them.

(Generative AI can also become Generative Thinking.)

Students learn to frame and reframe questions, approaching their subject matter from multiple angles simultaneously. They develop what some call "prompt literacy"—the ability to craft queries that open up possibilities rather than narrow them down. But more importantly, they learn to identify patterns and contradictions across AI responses, developing a meta-analytical skill that transcends any single interaction.

The 5 Questions Technique

• Forces you to clarify exactly what you want

• Helps the AI understand your context

• Creates a collaborative dialogue

• Improves final output quality

See it in action

I asked Claude (AKA: Kimmy/Claude) to research how AI might affect blue collar jobs 15-20 years in the future. I asked her to address jobs lost and what technologies would displace them. I added the “5 Questions” bit at the end.

Kimmy/Claude asked  the following questions. (Edited for brevity.) 

1. Are you particularly interested in any specific geographic regions or countries? 

2. Are there particular sectors you're most interested in (e.g., manufacturing, construction, transportation, maintenance, agriculture)?

3. Are you interested in how this might affect workers across different skill levels within blue collar work (entry-level vs. skilled trades)?

4. Would you like me to include potential retraining pathways or transition opportunities? 

5. Should I focus primarily on jobs that current blue collar workers could transition into, or should I take a broader view of entirely new industries and roles?

I answered: 

1. Northeastern US

2. Supply chain: warehouses, transportation and shipping 

3. All — particularly those with skills. 

4. YES, retraining and transition opportunities. 

5. New industries and roles.

Here’s The Thing

I’m writing a story set 15-20 years in the future. All I had in my head was that I wanted blue-collar workers to be part of the story. Thanks to those questions (and Claude’s replies), I’m now thinking about the implications of:

•  Where in the country those workers should be located.

• The physical jobs they do and how they’ll be replaced.

• What kind of “retraining” they might be given. 

• New jobs that would be created in a kind of hybrid “co-working” space.

All of which seem ripe for comedy. 

Pro tip

For deeper analysis, after you get your first reply, ask "Give me a Level 2 answer" - this tells the AI to provide more detailed, nuanced responses rather than basic summaries.

Your turn

What's your go-to prompting technique? Has the 5-question approach changed how you interact with AI? Share your experience in the comments.

I write about current AI thinking and practical advice on Mondays. On Wednesdays, I share my experiences writing the story currently known as Autonomous with Kimmy/Claude, my AI assistant. If you think you’d like to read and join in the discussion, please subscribe.

If you think you know someone who would enjoy diving into the topic of AI and Writing, please share this Substack with them.