## Core Concept
In the next 10 years (2026-2036), **agency is the only skill that makes all other skills valuable**. As AI commoditizes specialized knowledge, the ability to set your own direction and act without permission becomes the singular competitive advantage that determines who remains irreplaceable.
**Dan Koe's Thesis**: Agency is not just another skill—it's the **meta-skill** that determines whether your other skills matter at all.
## Why Agency Beats Specialization in AI Age
### The Commoditization Curve
```
Specialized Knowledge (AI can replicate)
↓
Becomes Cheap/Free
↓
Specialists Without Agency → Replaceable
↓
Generalists With Agency → Irreplaceable
```
**The Distinction**:
- **Specialist mindset**: "I have skill X, what jobs need skill X?"
- **Agency mindset**: "I have goal Y, what skills do I need to learn?"
### AI as Amplifier vs. Threat
**Low Agency + AI** = Displacement
- Waits for AI to be "good enough" to replace them
- Specializes deeper in shrinking domain
- Asks permission to use new tools
**High Agency + AI** = Multiplication
- Experiments with AI immediately
- Uses AI to expand into adjacent domains
- Treats AI as leverage, not competition
## The Experimental Mindset
### Life as Laboratory
High-agency people treat existence as ongoing experiment:
1. **Set direction** → Form hypothesis about what works
2. **Take action** → Run experiment without waiting for perfect information
3. **Gather data** → Observe results without emotional attachment
4. **Iterate** → Adjust based on evidence, not comfort
**Key Reframe**: Failure is not setback—it's data. Each failed experiment eliminates one path and provides information for the next iteration.
### The Permission Problem
Most people wait for:
- Job descriptions to match their skills
- Employers to recognize their value
- Market to demand their expertise
- Society to validate their direction
**High agency reversal**: Create the role, demonstrate the value, build the expertise, ignore the validation. Permission is not coming—and not needed.
## Cross-Domain Applications
### Career Strategy
**Traditional path** (low agency):
1. Develop specialized skill
2. Wait for company to hire you
3. Hope automation doesn't replace you
**Agency path** (high agency):
1. Identify problem worth solving
2. Learn whatever skills needed
3. Build solution publicly
4. Become irreplaceable through unique combination
### Learning Philosophy
**Low agency learner**: "What should I study to be employable?"
**High agency learner**: "What do I need to know to build X?"
The question you ask determines your trajectory.
### Entrepreneurship
**Low agency**: Find gap in market, build product, hope people buy
**High agency**: Build in public, iterate based on feedback, create demand through demonstration
The difference: waiting for validation vs. generating it through action.
### Personal Development
**Low agency**: Read books → Feel motivated → Do nothing → Repeat
**High agency**: Read books → Extract principles → Run experiments → Adjust → Build capability
Knowledge without experimentation is entertainment.
## The 10-Year Timeline
### 2026-2028: Early AI Disruption
- Specialized knowledge work becomes assisted
- High-agency workers use AI as leverage
- Low-agency workers resist and fall behind
### 2028-2032: Mass Commoditization
- AI handles most rote specialized tasks
- Generalists with vision become premium
- Specialists without agency struggle
### 2032-2036: Agency as Sole Differentiator
- Technical skills are table stakes (AI provides them)
- Direction-setting becomes the valuable skill
- High-agency generalists command premium
**Dan Koe's Warning**: The next 10 years is the transition period. Learn agency NOW while you can still leverage human advantages.
## How to Develop Agency
### 1. Start Small Experiments
Don't wait for perfect plan—run tiny experiments today:
- Write publicly about what you're learning
- Build something small without asking permission
- Reach out to someone "too important" to talk to you
- Learn one skill outside your domain this month
### 2. Treat Failure as Data
Reframe: "That didn't work" → "Now I know that path doesn't lead there"
**High agency question**: "What did this experiment teach me?"
**Low agency question**: "Why did I fail?"
### 3. Pursue Difficult Goals
Easy goals don't build agency—they reinforce comfort-seeking.
**Agency-building goals**:
- Have no clear playbook
- Require learning new skills
- Force uncomfortable conversations
- Make you look foolish to observers
The difficulty is the point. Agency grows through calibrated challenge.
### 4. Become Comfortable With Unclear Outcomes
**Low agency**: "I can't start until I know it will work"
**High agency**: "I'll start and adjust based on what happens"
Certainty is not available. Action under uncertainty is the skill.
## Related Frameworks
### [[High Agency Tricycle Model]]
Dan Koe's framework builds on the Tricycle Model:
- **Clear Thinking** → Identify direction worth pursuing
- **Bias to Action** → Run experiments without permission
- **Disagreeability** → Ignore social pressure to conform
Agency as meta-skill integrates all three.
### [[Generalist Advantage Framework]]
Generalists with agency adapt; specialists without agency stagnate.
Dan Koe's contribution: Agency is what makes generalist approach work. Without agency, generalists are just "interested in everything" but accomplish nothing.
### [[Default Path vs Pathless Path]]
Agency enables pathless path navigation. Without agency, only the default path feels safe.
### [[Five Low Agency Traps]]
Dan Koe's framework helps avoid these traps:
1. **Vague Trap** → Experimentation provides clarity
2. **Attachment Trap** → Iteration prevents over-attachment
3. **Midwit Trap** → Action beats analysis paralysis
4. **Rumination Trap** → Experiments generate data that ends loops
5. **Social Norms Trap** → Permission-free action breaks conformity
## Integration with Vault Knowledge
### Second Brain Methodology
Agency determines what you DO with captured knowledge. Without agency, Second Brain becomes hoarding.
**Connection**: [[../../Knowledge Management & Second Brain Methodology/README|Second Brain]] provides the infrastructure; agency provides the action.
### Career Strategy
Agency makes career strategy executable. Plans without agency remain fantasies.
**Connection**: [[../Career Strategy/README|Career Strategy]] documents show WHAT to do; agency is HOW you actually do it.
### Future of Work
Dan Koe's framework addresses THE question: How to remain valuable as AI advances?
**Answer**: Develop agency. Everything else becomes commodity.
**Connection**: [[../Future of Work/README|Future of Work]] explores structural changes; agency is personal response.
## Practical Implementation
### Weekly Agency Practice
1. **Monday**: Set one experimental goal for the week
2. **Daily**: Take one action without asking permission
3. **Friday**: Review what you learned from the week's experiments
4. **Document**: Write publicly about one insight
### Monthly Agency Audit
Ask yourself:
- What did I do this month that I wouldn't have done last year?
- How many experiments did I run?
- What did I learn from failures?
- Am I asking fewer permission questions?
**Agency grows through intentional practice, not passive accumulation.**
---
**Source**: [[../../Readwise/Documents/The most important skill to learn in the next 10 years]] - Dan Koe (4,570 words, January 2026)
**Created**: 2026-01-22
**Author**: Dan Koe framework synthesis
**Primary Domains**: Career strategy, learning methodology, AI adaptation, personal development
**Tags**: #agency #ai-age #meta-skill #experimentation #future-of-work #high-agency #dan-koe