Long-term Lock-in
Scenario
Long-term Lock-in
Model RoleDegradation Scenario
Primary DriversAI Ownership, Civilizational Competence
Sub-scenariosValues lock-in, Power lock-in, Epistemic lock-in
Related
Risk Factors
Scenarios
Overview
Section titled “Overview”Long-term Lock-in refers to scenarios where AI enables the permanent entrenchment of certain conditions - whether values, power structures, or epistemic environments. Unlike acute catastrophes, lock-in may unfold gradually and may not be immediately recognized as harmful.
Variants
Section titled “Variants”- Economic Power
- Political Power
- Epistemics
- Values
- Suffering Lock-in
Key Root Factors
Section titled “Key Root Factors”Loading diagram...
Factor Impact Scores
Section titled “Factor Impact Scores”85
↑ AI Ownership
Ownership patterns directly shape what gets locked in
80
↑ AI Uses
Where AI is deployed determines what structures get entrenched
75
↔ Civilizational Competence
High competence could enable good or bad lock-in
70
↑ AI Capabilities
More capable AI makes lock-in scenarios more feasible
60
↑ Misalignment Potential
Misaligned AI could lock in bad values or power structures
40
↑ Transition Turbulence
Crisis can entrench emergency powers and structures
35
↑ Misuse Potential
Misuse could trigger authoritarian responses
Dual Nature
Section titled “Dual Nature”Unlike other scenarios, lock-in is not inherently negative:
| Pole | Description |
|---|---|
| Positive | Beneficial values and institutions stabilized, protected from degradation |
| Negative | Authoritarian control, harmful ideologies, or stagnation permanently entrenched |
Outcomes
Section titled “Outcomes”Long-term Lock-in primarily affects:
- Long-term Trajectory: The quality and character of the post-transition future
- May contribute to Existential Catastrophe if lock-in prevents course correction