Curriculum

Cluster 2 ¡ Lesson 5 4 min read

Refining AI-Generated Content

Improve your auto-generated curriculum over time with targeted edits and your unique perspective.

The Refinement Mindset

You now have a complete curriculum. It's live. People can use it. And it's not perfect.

This is exactly where you should be.

The refinement phase is where your curriculum becomes genuinely valuable. AI gave you structure and competent explanations. You add the perspective, judgment, and voice that make it worth reading.

Prioritizing What to Fix

The Impact Matrix

Lesson Type Refinement Priority Why
Cluster 1, Lesson 1 Highest First impression, sets expectations
Foundation lessons High Vocabulary established here carries forward
Most-visited lessons High Actual user value
Dense/complex topics Medium Generic explanations hurt most here
Later specialized lessons Lower Readers who get here are already invested

Don't refine in order. Refine by impact.

Quick Wins First

Some fixes take 2 minutes and significantly improve quality:

  • Fix broken links: Nothing kills credibility faster
  • Add a specific example: Replace "for instance, in many fields..." with a real case
  • Cut hedging: "It might be argued" → just state the point
  • Add a sentence of context: "This matters because..."

Do these first. Save deep rewrites for later.

Adding Your Perspective

The Personal Insert

For each lesson, add at least one moment of genuine perspective:

Before (AI-generated):

"Berger and Luckmann's concept of institutionalization has been influential in sociology and related fields."

After (with your voice):

"Berger and Luckmann's concept of institutionalization changed how I think about organizational culture. When I see a company's 'way we do things,' I now ask: what repeated actions became habits, then rules, then 'just how things are'? The process is invisible until you have vocabulary for it."

The addition is small but transforms the lesson from summary to insight.

Where to Add Voice

  • Introduction: Why you chose this reading
  • Key concepts: What you think is most important
  • Assignment instructions: How you recommend approaching it
  • Knowledge check hints: What you found helpful to focus on

You don't need to add personal perspective everywhere. One or two moments per lesson is enough to make it feel authored, not generated.

Common Refinements

Generic → Specific

Before: "This concept appears in many contexts."

After: "This concept explains why design systems feel constraining at first but liberating once internalized—the initial friction is institutionalization in progress."

Abstract → Concrete

Before: "The author argues that social facts constrain individual behavior."

After: "Think about money. A $20 bill is paper. Its value exists only because we collectively agree it does. You can't individually decide it's worth $100. That's a social fact constraining your behavior—and you barely notice it."

Passive → Active

Before: "It has been suggested that media shapes perception."

After: "Lippmann argues that we don't react to the world—we react to 'pictures in our heads' that media creates. The news doesn't inform you about the world; it constructs the world you think you know."

The Weekly Refinement Practice

Build a sustainable habit:

Monday: Review One Lesson (15 min)

  • Read through as if you're a new learner
  • Note anything that confuses or feels flat
  • Don't fix yet—just note

Wednesday: Make One Improvement (20 min)

  • Pick the most impactful issue from Monday
  • Rewrite that section
  • Save and publish

Friday: Verify (5 min)

  • Check that your change deployed
  • Click through any links you touched
  • Done for the week

This cadence means 52 improvements per year. After a year, your curriculum is genuinely refined—and you never burned out doing it.

When to Stop Refining

Some lessons will never feel perfect. That's fine.

Stop refining a lesson when:

  • You've added your perspective to key sections
  • All links work and citations are verified
  • The explanation is clear to your target audience
  • You've read through it twice without finding issues

Move on. The remaining imperfections are less costly than the refinements you could make elsewhere.

The Living Curriculum

A curriculum isn't a document you finish. It's a practice you maintain.

As you learn more, add it. As readers give feedback, respond. As the field evolves, update. The AI gave you a starting point. Everything from here is yours.

Assignment

Begin refining your auto-generated curriculum:

Week 1: Foundation Lessons

  • Review the first lesson of each cluster
  • Add one personal insight to each
  • Fix any broken links or errors

Week 2: Key Concepts

  • Read through all key_concepts explanations
  • Rewrite any that feel generic
  • Add examples from your own experience

Week 3: Knowledge Checks

  • Test each question: Can it be answered without doing the reading?
  • Rewrite questions that are too easy or too vague
  • Add hints that guide without giving away answers

Ongoing: Response to Use

  • As you or others work through the curriculum, note friction points
  • Fix one thing per week
  • Let the curriculum evolve

The 80/20 of Refinement

Not all lessons need equal attention. Prioritize refinement based on:

  • Traffic: Which lessons do people actually read?
  • Foundations: Early lessons shape understanding of later ones
  • Complexity: Dense topics need clearer explanations
  • Your expertise: Where do you have unique insight?

Perfect the lessons that matter most. Leave others "good enough."

Adding Your Voice

AI-generated content is competent but generic. What's missing is you:

  • Your perspective: What do you think about this reading?
  • Your experience: How has this idea shown up in your work?
  • Your connections: How does this relate to other things you know?
  • Your emphasis: What matters most that others might miss?

The curriculum becomes valuable when it reflects your unique understanding.

Iterative Improvement

Don't try to perfect everything at once. Instead:

  1. Publish first: Get the draft live
  2. Use it: Work through your own curriculum
  3. Notice friction: Where do you stumble?
  4. Fix incrementally: One improvement per session
  5. Repeat: Ongoing refinement, not one-time editing

A curriculum that improves over months beats one that never launches.

Common AI Content Issues

Watch for these patterns in generated content:

  • Hedging language: "It could be argued that..." → State your view
  • Generic examples: Make them specific to your domain
  • Missing context: Why does this matter for your audience?
  • Superficial connections: Deepen links between readings
  • Hallucinated citations: Always verify sources exist

Each fix makes the curriculum more authentically yours.

Why publish imperfect content before refining it?

Hint: Think about feedback loops and motivation.

What makes your perspective valuable in a curriculum that AI helped create?

Hint: Consider what AI cannot know about your experience and judgment.

Working With Content Cluster

More on editing workflows using the CMS and local tools.

As soon as something becomes automated, so too does it become artisanal.