The AI Transition Model is a causal framework for understanding how various factors influence the trajectory of AI development and its ultimate outcomes for humanity. The interactive diagram above shows how Root Factors flow through Ultimate Scenarios to determine Ultimate Outcomes.

This model helps identify:

Leverage points: Which factors have the most influence on outcomes
Intervention targets: Where effort can most effectively shift trajectories
Key uncertainties: Which causal relationships are most uncertain
Scenario dependencies: How different pathways interact

How Parameters, Risks, and Interventions Connect

Both risks and interventions connect to root factors:

Risks (like deceptive alignment, racing dynamics) tend to increase harmful factors or decrease protective ones
Interventions (like interpretability research, compute governance) work to counteract risks

Interactive Views:

Parameter Table - Sortable tables with ratings (changeability, uncertainty, x-risk impact, trajectory)
Graph View - Visual causal diagram showing relationships between factors, scenarios, and outcomes
Data View - Raw data exploration interface

Why This Framing Matters

Traditional Risk Framing

“Trust erosion is a risk we must prevent”
“Concentration of power threatens democracy”
Focus: Avoiding negative outcomes

Parameter Framing

“Trust is a parameter that AI affects in both directions”
“Power distribution is a variable we can influence through policy”
Focus: Understanding dynamics and identifying intervention points

The parameter framing enables:

Better modeling: Can estimate current levels, trends, and intervention effects
Clearer priorities: Which parameters matter most for good outcomes?
Strategic allocation: Where should resources go to maintain critical parameters?
Progress tracking: Are our interventions actually improving parameter levels?

Relationship to Other Sections

Section	Relationship to Parameters
Risks	Many risks describe decreases in parameters (e.g., “trust erosion” = trust declining)
Interventions	Interventions aim to increase or stabilize parameters
Metrics	Metrics are concrete measurements of parameter levels
Models	Analytical models often estimate parameter dynamics and trajectories

How to Use This Section

For Researchers

Understand which underlying variables matter for AI outcomes
Identify gaps between current and optimal parameter levels
Design studies to measure parameter changes

For Policymakers

Prioritize interventions based on which parameters are most degraded
Monitor parameter trends to assess policy effectiveness
Coordinate across domains (a single parameter may affect multiple risks)

For Forecasters

Use parameters as input variables for scenario modeling
Estimate how different interventions would shift parameter levels
Identify tipping points where parameter degradation becomes irreversible