Skip to main content

Capabilities

When AI connects dots better than humans — what's left to connect?

Prediction

By 2027, AI will pattern match better than most knowledge workers. The capability that remains irreplaceable isn't connecting information. It's connecting with people.

AI ConnectsHumans Connect
DataPeople
PatternsMeaning
InformationConsciousness
DotsHearts

It's not what you know, but who you know you can trust that matters more than ever before.

Capability vs Skill

A skill is what you do. A capability is what enables you to do it.

CapabilitySkill
ScopeBroad, transferableSpecific, contextual
DurabilityDecadesYears (tools change)
AI impactAmplifiedOften automated
ExamplePattern RecognitionSQL data analysis
ExamplePromptingMidjourney v7 syntax
ExampleSellingCold calling

Skills get commoditised as AI absorbs them. Capabilities appreciate—they direct how skills are applied.

This matters for biological and phygital agents working together: humans bring capabilities that can't be replicated—judgment, taste, presence, trust. AI brings skills that deploy at scale—speed, consistency, volume, recall. The combination is greater than either alone.

The T-Shape

━━━━━━━━━━━━━━━━━━━━━  ← HORIZONTAL: Distribute transformative messages

┃ ← VERTICAL: Deep knowledge across domains

Wide enough to reach people. Deep enough to build what they need.

BarTypeExamplesWhat It Does
HorizontalSoftSelling, storytelling, negotiating, networking, writingGets value to people
VerticalHardCoding, engineering, designing, building, producingCreates value

Neither alone is enough. Build something nobody sees = waste. Sell something that doesn't exist = fraud.

The intersection is matrix thinking—extracting the meta from the matter, connecting domains AI treats as separate.

Beyond the T

The T-shape grows into a tree:

DimensionKnowsCan DoBuilds
DomainIndustry, customers, problemsTalk to users, identify JTBDEmpathy
BuildSoftware, systems, AI toolsShip product, automate, iterateExecution
ValueBusiness models, unit economicsPrice, measure, prove ROIJudgment
StrategyPositioning, competitionPrioritize, say no, find leverageFocus

Knowledge Types

TypeWhere It LivesCan AI Copy It?
Know-howIn your headHard — requires experience
CodifiedIn systems, manuals, codeEasy — AI excels here
EmbodiedIn practice, relationshipsHard — compounds over time

Most education focuses on codified. Most value comes from know-how. Most leverage comes from embodied.

The Agency Formula

Agency = Knowledge x Understanding x Capability x Intention x Execution

Without...You Are...
KnowledgeBlind
UnderstandingMechanical
CapabilityJust wishing
IntentionJust skilled
ExecutionJust planning

Most ideas die not because they're bad, but because the idea-haver stops too early in the chain.

Building Capability

1. Hit a wall    → Find something you can't do but need to

2. Copy blindly → Do exactly what someone successful does

3. Break things → Modify until it breaks, understand why

4. Fix things → Make it work again, but differently

5. Build things → Create from scratch using what you learned

6. Teach things → Transfer capability to others

Steps 2-5 are where capability lives. Most tutorials stop at step 2.

Credibility Chain

ConceptWhat Is It?How Earned?Role
CapabilitiesWhat you can doDemonstrated performanceFoundation
CredibilityBelief others haveReputation, outcomesMultiplier
CredentialsOfficial proofCertificates, licensesGatekeeper
CreditTrust extendedPast reliabilityLever

What if we hired based on what people built, not where they studied?

By Archetype

ArchetypeTop Capabilities
DreamerCreativity, Visualisation, Storytelling, Optimism, Selling
EngineerSystems Thinking, Planning, Pattern Recognition, Memory, Writing
RealistCritical Thinking, Listening, Pattern Recognition, Negotiation, Presenting
CoachQuestioning, Listening, Empathy, Presenting, Negotiation
PhilosopherFirst Principles, Critical Thinking, Questioning, Writing, Pattern Recognition

Go Deeper

CategoryCapabilityWhat It Does
ThinkingSystems ThinkingSee how parts connect
First PrinciplesReason from fundamentals
Critical ThinkingEvaluate without bias
Pattern RecognitionSee what others miss
PlanningTurn vision into steps
CommunicationStorytellingNarrative that moves
WritingClarity in text
PresentingPresence and delivery
ListeningUnderstand before responding
QuestioningUnlock better answers
PromptingDirect intelligence systems
ReadingAbsorb and synthesize
LanguagesConnect across cultures
ConnectionEmpathyFeel what others feel
SellingTrust transfer
NegotiationFind mutual value
OrchestrationWhen to think vs delegate
FoundationFocusDirect your scarcest resource
MemoryRetain what matters
TasteKnow good from good enough
CreativityGenerate options
VisualisationSee futures
OptimismPersist through setbacks

Tools Matrix

Which horizontal tools amplify which capabilities?

CapabilityToolAI %Strategy
StorytellingVideo Production60%Orchestrate
WritingWriting Tools70%Orchestrate
Systems ThinkingDiagrams30%Lead
SellingCRM40%Lead
PlanningProject Mgt50%Orchestrate
Pattern RecognitionAnalytics65%Orchestrate
PresentingPitch Deck45%Lead
QuestioningAI Interface35%Lead
PromptingLLMs, Image Gen25%Lead
FocusTime blocking, distraction blockers15%Lead
TasteCuration tools, quality benchmarks10%Lead

Lead = High human edge. Orchestrate = Direct AI effectively.

See Work Charts for how these shifts play out activity by activity.

The Human Edge

Human CoreAI AugmentsWhy It Matters
User empathy & tasteData synthesisAI can't feel the problem
Judgment under uncertaintyPattern recognitionAI needs clean signals
Trust & relationshipsResearch & analysisTrust requires skin in game
Ethical navigationCode generationAI has no stakes
Narrative & meaningTesting & iterationStories need a storyteller

Compound Effect

Single capabilities are commodities. Combinations are moats:

CombinationCreates
Writing + CodingDocumentation that actually helps
Design + EngineeringProducts people want to use
Sales + BuildingThings that actually sell
Teaching + DoingKnowledge that transfers
Domain + AISolutions others can't see

Distribution

Capability without distribution is worthless. The most valuable capability isn't technical—it's the ability to distribute what you build.

Mediocre builders with great distribution beat great builders with no distribution. Every. Single. Time.

The Curriculum

What should education actually teach?

Capability TypeHow to DevelopAI Can't Replace
ConnectingCoaching, masterminds, collaborationConsciousness meeting consciousness
ShippingProjects with deadlines, ship protocolJudgment under uncertainty
PredictingGames, sports, marketsIntuition from embodied experience
CharacterHeroes journey, friction, failureIntegrity under pressure

Traditional education inverts the priority—maximizing codified knowledge (AI's domain) while neglecting connecting and character (human's domain).

Practical Path

Stop collecting knowledge. Start building capability:

WeekAction
1Pick one thing to build (not learn—build)
2Build it publicly, let others see you fail
3Ship it broken, perfect is the enemy of shipped
4Fix it live, learn from real feedback
5+Build the next thing, each builds on the last

Where is your point of difference?

Context