Module 01 · Tier 4 — Intermediate Prompting

From User to Practitioner

The mindset shift that separates occasional users from people who consistently get results.

10 min Tier 4 — Intermediate Prompting

Why This Matters

There's a specific gap between people who use AI occasionally and people who use it well consistently. It's not a tool gap — they're using the same tools. It's a mindset gap. Occasional users treat AI as a shortcut. Practitioners treat it as a skill. This module is about making that shift explicit so you can accelerate through it deliberately.

The Concept

The Shortcut Mindset vs. The Practitioner Mindset

The shortcut mindset treats each AI interaction as a one-off transaction: you have a task, AI helps, you move on. No accumulation, no improvement loop, no system. Results are inconsistent because the inputs are inconsistent.

The practitioner mindset treats AI as a skill domain with its own learning curve. Each interaction is both output and data. What worked? What didn't? What would you do differently? Over time, you build intuition, documented patterns, and repeatable approaches that compound.

The practical difference: a practitioner finishes a task and spends 60 seconds noting what prompt approach worked. Six months later, that habit has built a personal knowledge base worth more than any prompt guide they could have bought.

The Four Shifts

1. From single prompts to conversation design

Occasional users write one prompt and accept what they get. Practitioners think in conversation arcs: how will I open this, what follow-ups will I need, how do I want to end the session? Designing the conversation before you start produces dramatically more useful outputs.

2. From hoping to evaluating

Occasional users read AI output and think "is this good enough?" Practitioners evaluate against explicit criteria: accurate, complete, appropriate tone, right length, serving the actual goal? The shift from passive reception to active evaluation changes what you accept and what you push back on.

3. From generic to specific

Occasional users use generic prompts that could apply to anyone. Practitioners build context into every prompt: their industry, their audience, their constraints, their voice. The more specific your context, the less the model falls back on generic training-data patterns.

4. From one-off to repeatable

The highest-leverage practitioner habit: when a prompt or approach works, document it. Not in a elaborate system — a simple note is enough. Over time, you build a prompt library that is specific to your work, your voice, and your real use cases. This is an asset that compounds.

The Practitioner Loop

Practitioners operate a continuous improvement loop on their AI use:

Intention: What am I trying to achieve? What does "done" look like?
Execution: Design the prompt with role, context, and format. Execute.
Evaluation: Did this hit the mark? Why or why not?
Capture: If it worked, note the approach. If it didn't, note what to change.

This loop takes 2 extra minutes per significant AI task. Over a year, it's the difference between stagnating at your current skill level and systematically improving.

The evaluation difference in practice

Here is the same output evaluated through two different lenses.

Output: An AI-generated executive summary of a 20-page report.

Shortcut evaluation: "That's pretty good. I'll use it." → paste and send.

Practitioner evaluation: Does this accurately represent the report's main argument? Are the three most critical findings here? Is the length right for this audience? Is the tone appropriate for an executive who didn't commission this work? Does it end with a clear implication or recommendation?

The practitioner catches three things the shortcut approach wouldn't: a key finding is missing, the third paragraph hedges too much for an executive audience, and the summary ends on data rather than implication.

Two minutes of structured evaluation versus two seconds of "looks fine." The output quality difference is significant.

Hands-On Exercise

Run the practitioner loop on a real task

ClaudeChatGPTGemini

Choose a task you would normally use AI for. Before you open any AI tool, write down: 1. What is the specific outcome I need? (Be precise — not "a good email" but "a 200-word email to a new client that establishes credibility without overselling") 2. What prompt approach will I use? (Role? Context? Format?) 3. What are my evaluation criteria? (What makes this output "done"?) Then execute the task and evaluate the output against your criteria. Did it hit all three? If not, what specifically missed? Finally: note one thing about your prompting approach that you'd change next time.

The pre-work is the point. Most people don't define what 'good' looks like before they start. That's why 'good enough' becomes the default.

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 are the four mindset shifts that separate occasional AI users from practitioners? Describe each one in your own words.

2 What is the practitioner loop, and why does the capture step matter?

Reflection

Honestly: which of the four mindset shifts do you most need to make? What's one concrete thing you'd do differently in your next AI session if you were operating as a practitioner rather than an occasional user?

Key Takeaway

The gap between inconsistent and consistently excellent AI results is a mindset gap, not a tool gap. Practitioners evaluate deliberately, build context into every prompt, and capture what works. The practitioner loop compounds — occasional use doesn't.