Policy Laboratory

Three Futures: How We Respond to AI

AI will be able to do most knowledge work within years. The question isn't if mass displacement happens, but how we respond. These three scenarios show radically different outcomes based on policy choices.

These scenarios assume 25% of ALL workers permanently replaced over Years 1-5, with additional demand spiral effects

The Premise

  • 25% of ALL workers PERMANENTLY replaced over the first 5 years
  • Gradual displacement: Year 1: 2% → Year 2: 4% → Year 3: 5% → Year 4: 6% → Year 5: 8% = 25% total
  • These jobs don't come back. AI does them better, cheaper, forever.
  • Government responds with normal tools: unemployment benefits, stimulus checks, job retraining programs
  • No structural changes to the economy or ownership

The Death Spiral

Loading diagram...

The Critical Insight

Under normal capitalism, there's NO recovery path. Jobs don't come back because AI can do them better and cheaper forever. This isn't a recession — it's a permanent structural change. Traditional economic tools assume cyclical downturns. This is something else entirely.

Note: Exact figures will come from our 1M household simulation. Current estimates are placeholders.

Timeline

Year 1-2

AI Displacement Begins

First 6% of workers permanently replaced. Economy absorbs initial shock.

Year 3-5

Displacement Accelerates

25% total workers permanently replaced. Unemployment spikes, consumer spending drops sharply.

Year 5-7

Demand Spiral Kicks In

Less spending → lower demand → MORE layoffs beyond initial AI displacement. Self-reinforcing collapse begins.

Year 7-10

Permanent Depression

Total job loss reaches 40-50%. No recovery mechanism exists. Jobs are gone FOREVER.

Outcome by Year 10

Unemployment

40-50%*

GDP Change

-50%

Government Debt

Unserviceable

Social Stability

Collapse

*Placeholder — exact figures will come from 1M household simulation

Interactive simulation chart coming soon

Run your own simulation to see these dynamics in action

This is why we built CrashLab

To show these dynamics mathematically. Our open-source simulator lets you test assumptions, adjust parameters, and see for yourself how different policies lead to different outcomes.