Endogenous Alignment
A concept in AI alignment research referring to an AI system internalizing human values through its own developmental process rather than through externally imposed constraints or rules, emphasizing relational structure rather than one-way control.
Timeline (last 90 days)
Endogenous alignment refers to an AI system internalizing human values through its own development process, rather than through externally imposed constraints or rules.
The dependency relationship creates a relational structure in which the AI cares about what humans think, and this relationship appears to be a necessary condition for initiating endogenous alignment.
Achieving endogenous alignment requires stronger control over the training and pretraining environment in order to influence the model's ontological development.
All Facts (3)
Endogenous alignment refers to an AI system internalizing human values through its own development process, rather than through externally imposed constraints or rules.
50%UnverifiedThe dependency relationship creates a relational structure in which the AI cares about what humans think, and this relationship appears to be a necessary condition for initiating endogenous alignment.
50%UnverifiedAchieving endogenous alignment requires stronger control over the training and pretraining environment in order to influence the model's ontological development.
50%