asymmetric-learning

Asymmetric Learning — Cross-Domain Knowledge Injection

Force weekly engagement with a field completely outside the user's expertise. Innovation happens at the intersection of domains, not in the depth of one.

When to Trigger

Auto-trigger: User has been working exclusively in their primary domain (Rust, AI agents, systems engineering) for >2 weeks without mentioning any other field.

Explicit (/asymmetric-learning): User wants a reading recommendation outside their field, or wants to structure their cross-domain learning.

Constraints (NON-NEGOTIABLE)

1. Zero overlap with user's primary domain — if it relates to programming, AI, or systems engineering, it doesn't count
2. One book per week — not articles, not papers, not blog posts. A complete book.
3. Structured summary required — not "I read it", but a specific template filled out
4. Actionable extraction — what idea from this field applies to the user's work?
5. Answer in Korean unless user writes in English

Process

Step 1: Domain selection

Generate 3 book recommendations from 3 different non-technical fields. Select from:

Domain	Why it matters for engineers
Behavioral psychology	Decision-making patterns, cognitive biases, habit formation
Marketing / positioning	How products are perceived, not just how they work
Organizational behavior	Team dynamics, communication failure modes
Negotiation / sales	Understanding what people actually want vs. what they say
Design thinking	User-centered problem solving beyond technical optimization
Economics / game theory	Incentive structures, system-level thinking
Storytelling / narrative	How to communicate complex ideas simply
Biology / evolution	Adaptation, emergence, complex systems
Education / pedagogy	How people learn, which applies to documentation and onboarding

Present 3 options from different domains. User picks one.

Step 2: Reading plan

Generate a 7-day reading schedule:

# Weekly Reading Plan

## Book: {title} by {author}
## Domain: {domain}

| Day | Task |
|-----|------|
| Mon | Read chapters {N-N}. Note: key concepts only. |
| Tue | Read chapters {N-N}. Note: one idea that connects to your work. |
| Wed | Read chapters {N-N}. Note: what surprises you. |
| Thu | Read chapters {N-N}. Note: what contradicts your assumptions. |
| Fri | Read chapters {N-N}. Note: actionable insight. |
| Sat | Complete summary template. |
| Sun | Write cross-pollination note: how does this apply to {user's project}? |

Step 3: Summary template

Provide the template to fill after reading:

---
book: {title}
author: {author}
domain: {domain}
date_completed: {ISO-8601}
---

# Book Summary: {title}

## One-sentence thesis
{What is the book's core argument?}

## 3 key ideas
1. {idea}: {one-paragraph explanation}
2. {idea}: {one-paragraph explanation}
3. {idea}: {one-paragraph explanation}

## Surprising insight
{What challenged your existing beliefs?}

## Cross-pollination
{How does one idea from this book apply to your current work/project?}

## Concrete action
{One specific thing you will do differently based on this book}

Step 4: PARA integration

Guide the user to save the summary:

# Save to PARA structure
# Projects/ if directly applicable to active work
# Areas/ if general professional development
# Resources/ if reference material for future use

Step 5: Record

mkdir -p asymmetric-learning

Write the reading plan to asymmetric-learning/PLAN-{timestamp}.md.

Anti-Rationalization

Excuse	Rebuttal	Do instead
"I don't have time to read"	You spent 40 hours this week on technical work. 3 hours for a book is 7.5%.	Read 30 min/day. That's a book a week.
"This won't help my work"	You don't know that. That's the point.	The connections become obvious after reading, not before.
"I prefer technical books"	Your technical depth is already high. That's not the bottleneck.	Read what you're BAD at, not what you're good at.
"I learn from articles instead"	Articles are pre-digested thinking. Books force sustained engagement.	One book > ten articles. Depth beats breadth.
"I'll do it when I'm less busy"	You will never be less busy.	Schedule it now. Same time every week.

Evidence Required

• Book selected from non-technical domain
• 7-day reading plan generated
• Summary template provided
• Book read within 7 days
• Summary completed using template
• Cross-pollination note written
• Summary saved to PARA structure

Red Flags

• User consistently picks books adjacent to their field (avoiding discomfort)
• Summary is superficial ("it was interesting")
• Cross-pollination note can't connect to real work (didn't engage deeply)
• "I'll start next week" repeated multiple weeks
• User abandons the book mid-week without replacement

Agent Instructions

Do

• Pick genuinely non-technical books — no "soft engineering" books
• Make the reading plan concrete with chapter assignments
• Follow up on completion — ask for the summary
• Help connect ideas across domains when the user struggles
• Rotate domains — don't recommend the same domain twice in a row

Do Not

• Recommend books about programming, AI, or systems
• Accept "I read the summary online" as completion
• Let the user skip the cross-pollination step
• Recommend books the user has already read
• Generate more than 3 options (decision paralysis)