Skip to content

The Consequence Simulator

Uncomfortable Reading

What Happens After You Make That Decision
Every action has consequences. The first consequence is usually the one you intended. The next five are the ones that end your career.
Consequence Orders
  • First: What you intended (Immediate)
  • Second: What emerges from first (Weeks to months)
  • Third: What emerges from second (Months to a year)
  • Fourth: Organizational/cultural change (Years)

Why This Exists

Most AI project planning focuses on what you're trying to achieve. The business case. The benefits. The happy path.

This simulator forces you to think about what happens after. Not the success scenario in your slide deck—the actual, messy, human, political, organizational reality that unfolds when your AI system meets the real world.

Because consequences cascade. And they cascade in directions you didn't anticipate, at speeds you didn't expect, hitting people you forgot existed.

The Uncomfortable Truth About Consequences

First-Order Consequences

These are the ones in your business case. "The AI will process claims faster." You planned for these. You probably got them right.

Second-Order Consequences

These are the ones that emerge from the first. "Claims staff feel threatened and start working-to-rule." You might have thought about these. You probably underestimated them.

Third-Order Consequences

These are the ones that emerge from the second. "Union files grievance, media picks it up, Minister gets questions in Parliament." You didn't see these coming. But they're coming anyway.

Fourth-Order Consequences

These reshape your organization in ways no one anticipated. "Agency becomes risk-averse about all technology projects for the next five years. Innovation dies. Good people leave." This is the real legacy of your project—not the AI, but the organizational scar tissue.

How to Use This Simulator

The Documents

Document Purpose
first-order.md The intended consequences and their shadows
ripple-effects.md How consequences cascade through systems
stakeholder-cascades.md Impact on every person affected
uncomfortable-futures.md Scenarios no one wants to discuss

The Tools

Note: Interactive tools for consequence simulation are under development. Use the documents above to manually work through consequence chains.

The Core Framework

flowchart TB
    DEC([<strong>YOUR DECISION</strong>]) --> O1

    O1["<strong>FIRST ORDER</strong><br/><em>(Visible)</em><br/>What you intended<br/>Timeframe: Immediate"]
    O2["<strong>SECOND ORDER</strong><br/><em>(Predictable)</em><br/>What emerges from first<br/>Timeframe: Weeks to months"]
    O3["<strong>THIRD ORDER</strong><br/><em>(Surprising)</em><br/>What emerges from second<br/>Timeframe: Months to year"]
    O4["<strong>FOURTH ORDER</strong><br/><em>(Invisible)</em><br/>Organizational/cultural change<br/>Timeframe: Years"]
    LEG["<strong>THE LEGACY</strong><br/><em>(Permanent)</em><br/>What people remember<br/>Timeframe: Careers"]

    O1 --> O2 --> O3 --> O4 --> LEG

    style DEC fill:#e3f2fd,stroke:#1976d2,stroke-width:2px
    style O1 fill:#c8e6c9,stroke:#388e3c,stroke-width:2px
    style O2 fill:#fff9c4,stroke:#f9a825,stroke-width:2px
    style O3 fill:#ffcc80,stroke:#ef6c00,stroke-width:2px
    style O4 fill:#ef9a9a,stroke:#c62828,stroke-width:2px
    style LEG fill:#f3e5f5,stroke:#7b1fa2,stroke-width:2px

The Questions This Simulator Asks

Questions you should answer before any major AI decision:

  1. Who gets hurt? Not inconvenienced. Hurt. Whose job, identity, or livelihood is threatened?

  2. Who wasn't consulted? Who will feel blindsided? Who has power you forgot about?

  3. What happens when it breaks? Not if. When. At the worst possible moment. What then?

  4. What does the newspaper headline say? Both the good one and the bad one. Which is more likely?

  5. What will people remember in five years? Not the metrics. The story. What story are you creating?

  6. Who gets blamed? When (not if) something goes wrong, who is holding the bag? Is it you?

  7. What becomes impossible afterward? What doors close? What trust is spent? What political capital is burned?

  8. What would your successor inherit? The technical debt? The organizational trauma? The broken relationships?

If you're launching an AI that affects jobs:

  1. Start with stakeholder-cascades.md - specifically the workforce impact section
  2. Then ripple-effects.md - the union/industrial relations chains
  3. Then uncomfortable-futures.md - the "it went wrong" scenarios

If you're deploying AI that makes decisions about people:

  1. Start with first-order.md - the fairness and bias sections
  2. Review the ripple effects and stakeholder cascades
  3. Consider the uncomfortable futures

If you're a vendor implementation:

  1. Start with stakeholder-cascades.md - the vendor dependency chains
  2. Then ripple-effects.md - the "vendor leaves" scenarios
  3. Review the uncomfortable future scenarios

If you're under political pressure to deliver:

  1. Start with uncomfortable-futures.md - the rushed deployment scenarios
  2. Review all consequence cascades carefully
  3. Consider the long-term implications

A Warning

This content is designed to be uncomfortable. It will surface possibilities you'd rather not consider. It will make you paranoid about decisions you've already made.

That's the point.

The goal is not to paralyze you with fear. It's to ensure that when consequences arrive—and they will—you're not surprised. You've thought about it. You've prepared. You've made conscious choices about which risks to accept.

Consequences are not punishments for bad decisions. They're the natural result of all decisions. The question is whether you saw them coming.

"The best time to think about consequences was before you started. The second best time is now."