Activation Energy

#core-framework #computational-lens

What It Is

Activation energy is the cost to initiate a behavior. Every action has an initial threshold you must breach before it becomes self-sustaining.

This isn't metaphorical. Your brain literally requires metabolic resources to initiate new patterns. Starting work costs 4-6 willpower units. Continuing work costs ~0.5 units.

Starting is the hardest part. Not because you're weak. Because physics.

The Mechanism

Different behaviors have different activation costs:

Low activation energy (< 1 unit):

  • Checking phone (if visible): ~0.5 units
  • Following established routine: ~0.5 units
  • Continuing current activity: ~0.5 units

Medium activation energy (2-4 units):

  • Making a decision: 2-3 units
  • Active resistance to temptation: 2-3 units
  • Context switching: 0.5-1 unit per switch

High activation energy (4-6 units):

  • Threshold breach (starting work from rest): 4-6 units
  • Cold-starting after dormancy: 6+ units
  • Initiating difficult conversations: 4-6 units

The behavior that runs is whichever has the lowest cost when the decision point arrives.

The 30x30 Pattern

Activation energy decreases with repetition over approximately 30 days.

The rough curve:

  • Days 1-7: High cost, forcing yourself (6 units)
  • Days 8-15: Cost decreasing (4 units)
  • Days 16-30: Approaching automatic (2 units)
  • Day 31+: Feels effortless (0.5 units)

This is based on habit formation research and observable in real practice. After 30 days of consistency, the neural pathways are cached and execution becomes automatic.

Why Starting Is Hardest

When you start a behavior, you're:

  1. Loading context into working memory - What am I doing? Why? What's the first step?
  2. Overcoming inertia - Transitioning from low-energy rest_state to high-energy work_state
  3. Building the execution pathway - Neural patterns not yet cached, must be computed fresh

After you've started:

  • Context is already loaded
  • Momentum is established
  • Pathways are active
  • Continuing is cheaper than stopping

This is hysteresis - path dependence. Same task, different starting state, massively different energy cost.

The Boltzmann Distribution

From statistical mechanics: systems naturally fall into low-energy configurations.

The probability of being in a state is proportional to e^(-E/kT) where E is the energy of that state.

Translation: You naturally do what's easiest. This isn't moral weakness. It's thermodynamics.

So-called disciplined people don't fight this law of physics. They engineer environments so desired behaviors have lower energy than undesired ones.

Example: Phone on Desk vs Drawer

graph TD
    subgraph "Phone on Desk (Bad Landscape)"
    A1[Work State] -->|Check phone: 0.1 units| B1[Phone State]
    A1 -->|Resist: 2 units each time| A1
    B1 --> B1
    style B1 fill:#ff9999
    end

    subgraph "Phone in Drawer (Engineered Landscape)"
    A2[Work State] -->|Check phone: 4 units| B2[Phone State]
    A2 -->|Continue work: 0.5 units| A2
    style A2 fill:#99ff99
    end

Phone on desk:

  • Energy to check: 0.1 units (just pick it up)
  • Energy to resist: 2 units (active inhibition every time you see it)
  • Boltzmann distribution says: you'll check the phone

Phone in locked drawer:

  • Energy to check: 4 units (stand, walk, unlock, retrieve)
  • Energy to continue working: 0.5 units (already in work state)
  • Boltzmann distribution says: you'll keep working

Same person. Different energy landscape. Different behavior. Not discipline. Thermodynamics.

Bridge Scripts

Cold starts are expensive. Bridge scripts heat up the system to reduce activation cost.

Instead of going from rest_state (0 energy) → work_state (6 units), you build a sequence:

  1. rest_state → wake_state (automatic, circadian)
  2. wake_state → alert_state (morning light, 1 unit)
  3. alert_state → ready_state (10-min braindump, 2 units)
  4. ready_state → work_state (pre-work checklist, 1 unit)

Total cost: 4 units distributed across 30 minutes. Much cheaper than trying to breach 6 units all at once.

By the time you sit down to work, work_state is now the lowest-energy option. The activation energy has already been paid through the bridge sequence.

Detraining and Reactivation

When you stop executing a pattern for months, neural pathways atrophy. Not moral failure. Physical degradation from disuse.

After 3-month dormancy, expect:

  • Week 1: 20% of previous capacity, high activation cost
  • Week 2: 40% capacity, cost decreasing
  • Week 3: 60% capacity, approaching pre-detraining state
  • Week 4+: 80%+ capacity, fully reactivated

This is detraining - neural pathways atrophying from disuse. Treat it like fitness. Start at 20%, build gradually through consistent execution.

Don't expect day-1 performance after long breaks. You're cold-booting a system that's been powered down. High activation energy is predictable and temporary.

Reducing Activation Energy

1. Make It Concrete

"Work on project" = ambiguous = high cognitive load at decision point

"Write 300 words of documentation for skills API" = specific = lower cognitive load

Ambiguity increases activation cost. Specificity lowers it.

2. Install Triggers

Time-based: "At 9 AM I work" Location-based: "When I enter library I work" Ritual-based: "After braindump I work"

Triggers eliminate the decision about when to execute. Reduces activation cost by removing one decision layer.

3. Remove Competing Options

Can't be tempted by phone if it's not accessible.

Can't choose Netflix if it's blocked.

Can't order delivery if apps are deleted.

Prevention costs 0 units. Resistance costs 2-3 units.

4. Rehearse Until Cached

After 20-30 repetitions of simple behaviors, they become automatic.

After 60-90 repetitions of complex behaviors, they become automatic.

The pattern caches. Activation cost drops to near-zero. Execution becomes effortless.

Why "Just Start" Fails

"Just start" ignores the activation energy cost.

You're telling someone to pay 6 willpower units with no strategy for making that payment affordable. If they only have 3 units available, the command is physically impossible.

Better advice: "Build a bridge script that distributes the 6-unit cost across a 30-minute warm-up sequence."

Backward Chaining Reduces Activation Cost

Backward chaining systematically lowers activation energy by working backward from the goal to identify concrete prerequisites. Forward planning asks "What should I do?" from unbounded current state and faces infinite branching—this is exponentially expensive and triggers analysis paralysis. Backward planning asks "What requires work launched?" from specific goal state and follows constrained prerequisite relationships.

The backward chain for work launch might be: Work launched requires sequence executed. Sequence executed requires script loaded. Script loaded requires morning mantra and braindump completed. Braindump requires waking on time. Waking on time requires sleep by 10pm. Each step backward adds specificity and reduces ambiguity. By the time you reach "sleep by 10pm," the action is concrete and the path forward is clear.

This approach reduces activation energy because it eliminates the cognitive work of figuring out "where to start." The goal provides automatic filtering that prunes irrelevant paths. Instead of staring at an overwhelming project wondering what to do first (6 units to breach that ambiguity), you execute a predetermined sequence where each step clearly follows from the previous (0.5 units per step after the sequence is cached).

Key Principle

Engineer for low activation, not high willpower - Reduce the energy cost to start desired behaviors through bridge scripts, triggers, and specificity. Increase the energy cost of undesired behaviors through prevention architecture.

Starting isn't hard because you're lazy. Starting is hard because it costs energy to breach the threshold. Build systems that make the energy cost manageable.