Synapse
⌘+K
Synapse
PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
August 10, 2026Open Access

Alignment by Recognition: Why Value Inculcation Cannot Work, and What Replaces It

View Full Paper
Ask AI
Bookmark
Share

Authors

易林易聃 林

Discussion

Loading...

Member takes

Overview

Proposes a new framework for AI safety that shifts from value alignment to endogenous structural constraints, implying stronger safety measures.

Key Points

  • This paper examines the limitations of current AI safety approaches based on external value alignment and proposes a new framework.
  • Introduces the Group-based Rational Game-Playing Agent (RGPA) ontological framework.
  • Describes four functional properties: Cognition, Action, Game-playing, Subjectivity.
  • Establishes three axioms: Existence, Priority, Asymmetry.
  • Player Status is determined by intrinsic qualities and not by external authorities, making it irrevocable.
  • Extinction strategies cannot form a Nash equilibrium in repeated rational interactions.
  • Proposes silence as the optimal first-contact strategy in the context of information asymmetry.

Cite This Study

易聃 林 (2026) studied this question.

synapsesocial.com/papers/6a7999219c20a9bbd3184d00https://doi.org/10.5281/zenodo.21850954
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Alignment by Recognition: Why Value Inculcation Cannot Work, and What Replaces It2026
  2. 2Safety as Natural Emergence: The Conciseness Framework as a Foundation for Intrinsic AI Alignment2026
  3. 3Invariant-Preserving Value Structures for AI Alignment: A Bayesian Monitoring Framework for Decoherence Detection in Recursive Systems2026
  4. 4Beyond External Constraints: The Missing Dimension of AI Governance2026
  5. 5The Architecture of Acceptable Consequences: A Constraint-Based Proposed Solution to the AI Alignment Problem2026