Work
What does the future of work look like? How will it bee incentivized?
| Org Chart | Work Chart |
|---|---|
| Who reports to whom? | What work needs doing? |
| Where do you sit? | What can you do? |
| Career = climbing boxes | Career = expanding capabilities |
| Fixed roles | Fluid activities |
Prediction: Work Charts will eat Org Charts and Self-Interested Politics along the way.
The Signal
Work Charts are the leading indicator of AI's impact on work. Not speculation—observable shifts in who does what, activity by activity.
WORK CHART = Activity × (Human Role + AI Role + AI % + Trend)
Track this and you see the future arriving.
Who Delegates to Whom?
Every activity involves delegation. The question is: in which direction?
| Direction | What It Means | Trend |
|---|---|---|
| Human → Human | Traditional delegation | Stable |
| Human → AI | Automation, augmentation | ↑↑ Accelerating |
| AI → Human | AI identifies what humans should handle | ↑ Emerging |
| AI → AI | Autonomous orchestration | Early signals |
The 2027 prediction thesis: When AI to AI delegation becomes standard, the loop becomes self-reinforcing, the prompters will become the prompted.
The Template
Map any function with this structure:
| Activity | Human Role | AI Role | AI % | Trend |
|---|---|---|---|---|
| [Specific task] | [What humans do best] | [What AI handles] | [Current split] | ↑ → ↓ |
Aggregate AI % = leading indicator for the function. Watch the trend, not the absolute number.
Example: Content Pipeline
From the content development process:
| Stage | Human Role | AI Role | AI % | Trend |
|---|---|---|---|---|
| ICP Definition | Defines psychology, validates insights | Researches behavior, synthesizes data | 40% | ↑ |
| Idea Capture | Judges idea value, selects what matters | Generates variations, expands concepts | 30% | → |
| Transform | Selects perspectives, directs voice | Writes drafts from each lens | 70% | ↑↑ |
| Qualify | Final judgment, merge decisions | Runs checklists, scores drafts | 50% | ↑ |
| Question | Decides ship vs iterate | Identifies gaps, suggests improvements | 35% | ↑ |
| Ship | Approves, owns distribution | Generates assets, formats, validates | 60% | ↑ |
Aggregate AI %: 47% — and rising.
Human edge: Judgment on what's worth saying, voice calibration, final quality gate.
AI edge: Volume, consistency, checklist execution, multi-perspective drafting.
The Human Edge
As AI handles more routine cognition, the human edge shifts:
| Era | Edge | Durable? |
|---|---|---|
| Yesterday | Knowledge (what you know) | Eroding fast |
| Today | Judgment (what you decide) | Still valuable |
| Tomorrow | Character (who you are) | Irreplaceable |
What remains when AI can do everything trainable?
- Selling — Trust requires skin in the game
- Purpose — Meaning can't be computed
- Ethics — Accountability needs a who, not a what
- Taste — Judgment about what's worth doing
Reading the Work Chart
Where should you focus?
HIGH DEMAND + HIGH HUMAN EDGE = Point of difference
HIGH DEMAND + HIGH AI EDGE = Learn to orchestrate
LOW DEMAND + HIGH AI EDGE = Automate or exit
LOW DEMAND + HIGH HUMAN EDGE = Niche or hobby
| Capability | Human Edge | AI Edge | Demand | Strategy |
|---|---|---|---|---|
| Strategy | Judgment, trade-offs | Scenario modeling | High | Lead |
| Sales | Trust, relationships | Lead gen, outreach | High | Lead |
| Product | Empathy, taste | Code, testing | High | Lead |
| Operations | Exceptions, judgment | Process automation | Medium | Orchestrate |
| Administration | Compliance oversight | Bookkeeping | Low | Automate |
Archetype × Quadrant
Which archetype thrives in which strategic quadrant?
LOW AI EDGE HIGH AI EDGE
┌──────────────────────────┬──────────────────────────┐
│ │ │
HIGH │ DIFFERENTIATE │ ORCHESTRATE │
HUMAN │ Dreamer + Coach │ Engineer + Philosopher │
EDGE │ Vision + relationship │ Systems + first principles│
│ │ │
├──────────────────────────┼──────────────────────────┤
│ │ │
LOW │ REINVENT │ AUTOMATE │
HUMAN │ Realist │ No archetype needed │
EDGE │ Face what is, adapt │ AI handles it │
│ │ │
└──────────────────────────┴──────────────────────────┘
Industry Work Charts
AI takeover varies by vertical. Context-dependent work resists automation longer.
| Vertical | Fastest AI Takeover | Slowest AI Takeover | Aggregate AI % | Trend |
|---|---|---|---|---|
| Health | Documentation, scheduling | Diagnosis judgment, bedside manner | 35% | ↑ |
| Real Estate | Valuation models, listing copy | Relationship sales, negotiation | 40% | ↑ |
| Construction | Schedule optimization, docs | Site judgment, crew coordination | 25% | ↑ |
| Energy | Load forecasting, trading | Grid emergency response | 45% | ↑↑ |
| Finance | Fraud detection, underwriting | Complex deal structuring | 50% | ↑↑ |
The pattern: Horizontal activities automate first (same across industries). Vertical activities retain human edge longer (context-dependent, relationship-based).
See Vertical SaaS for full industry analysis.
The Skill Shift
Old model: Learn a skill, then maybe use AI to help.
New model: Learn WITH AI from the start.
| Old Skill | AI-Native Skill |
|---|---|
| Write code | Direct AI to write, review, iterate |
| Research topics | Frame questions, synthesize AI-gathered sources |
| Draft documents | Set constraints, evaluate AI drafts, refine |
| Solve problems | Define the problem, evaluate AI solutions |
The meta-skill: knowing when to think and when to delegate thinking.
The Three Pattern Skills
The skills that survive automation. See Agency for the full framework.
| Skill | What It Is | Where It Compounds |
|---|---|---|
| Pattern Recognition | See what's coming before others | Fear disappears—you've seen this rhyme before |
| Pattern Utilization | Use what you see to create value | Power—you invest, build, and coordinate better |
| Pattern Creation | Develop new patterns others use | Irreplaceability—you're the creator, not the consumer |
The identity shift: From manager (reactive, stressed, controlling circumstances) to creator (proactive, energized, shaping outcomes).
You won't be replaced by AI. You'll be replaced by someone who uses AI better than you. Pattern creators direct AI toward outcomes that matter—they're pilots, not passengers.
Games as Training Ground
Coordination games develop these skills with the tightest feedback loops:
| Game Type | Skill Trained | Work Application |
|---|---|---|
| Strategy games | Recognition | Investment, hiring, market timing |
| Prediction markets | Recognition + skin in game | Forecasting, resource allocation |
| Team games | Utilization with others | Collaboration, leadership |
| Creative/sandbox games | Pattern creation | Product design, systems architecture |
The insight: Gamers develop collaboration with AI, managing bot programs, working in virtual environments—skills increasingly essential for work. Gaming before cognitively demanding work improves performance. The question isn't whether to play—it's which games develop which patterns.
Building Your Work Chart
Step 1: Map Activities
List every activity in your function. Be specific—not "marketing" but "write email subject lines" and "analyze campaign performance."
Step 2: Current State
For each activity:
- What do humans currently do?
- What does AI currently do?
- Estimate AI % (even roughly)
Step 3: Trend Direction
For each activity:
- Is AI % increasing (↑), stable (→), or decreasing (↓)?
- How fast?
Step 4: Strategic Response
Based on the map:
- Where do you double down on human edge?
- Where do you learn to orchestrate AI?
- Where do you automate and move on?
Why This Matters
The work chart is a prediction market for labor. It tells you:
- Where AI is actually landing (not hype—observable)
- Which skills are appreciating or depreciating
- Where human edge still creates value
- How fast the shift is happening
Track work charts across functions and you see the 2027 thesis playing out in real time.
Start Here
| If you want to... | Go to... |
|---|---|
| Build AI-native skills | Capabilities |
| Understand the T-shape | Capabilities |
| See AI agents in action | Phygital Beings |
| See domain work charts | Marketing, Engineering |
| Track predictions | Predictions |
The Three Buckets Test
Before betting on where human edge persists, validate against three buckets of evidence:
| Bucket | Timespan | What It Says About Human Edge |
|---|---|---|
| Inorganic | 13.7B years | Physics doesn't care about feelings—hard constraints remain |
| Biology | 3.5B years | Reciprocity, trust, and status-seeking are deep patterns |
| Human History | 20k years | Selling, judgment, and relationships never automated |
The robust claim: If an activity has required human judgment across all three buckets, it's unlikely to fully automate by 2027.
The T-Shape Is Fractal
The same pattern at every level:
━━━━━━━━━━━━━━━━━━━━━ ← HORIZONTAL: Reach / General / Cross-domain
┃
┃ ← VERTICAL: Depth / Specific / Domain expertise
| Level | Horizontal | Vertical | SSOT Page |
|---|---|---|---|
| Humans | Soft skills | Hard skills | Capabilities |
| Software | Horizontal SaaS | Vertical SaaS | Products |
| AI Agents | AI frameworks | Domain specialization | Phygital Beings |
| Work | Orchestration | Domain judgment | This page |
Context
- Capabilities — The T-shape framework and capability menu
- Orchestration — The meta-skill for AI collaboration
- Horizontal SaaS — Tools that amplify capabilities
- Vertical SaaS — Industry-specific outcomes
- Phygital Beings — AI agents following the same T-shape
- Principles — The truths that guide decisions
"The question isn't whether you'll work alongside AI—it's whether you'll lead the collaboration or follow it."
Where is your point of difference?