What question did this study set out to answer?

This research develops a structural theory to define ethical admissibility for autonomous agents without relying on optimization or utility.

April 6, 20260 citationsOpen Access

View Full Paper

Epistemic Safety Without Optimization: A Structural Theory of Ethical Admissibility for Autonomous Agents

DGDaniel Pineda González

Key Points

This research develops a structural theory to define ethical admissibility for autonomous agents without relying on optimization or utility.
Introduced the Lógica Ética para Sistemas Operativos (LETOS) framework.
Defined ethical action membership using structural constraints derived from axioms.
Proved various theoretical results including boundary characterization and authority separation.
Explored multi-agent coordination without merging knowledge.
Characterization of the admissible action region (Ω) was completed.
Established that agents do not optimize but check for membership inΩ.
Demonstrated the capability for coordination among multiple agents under the new framework.

Abstract

We present LETOS (Lógica Ética para Sistemas Operativos), an axiomatic structural theory for autonomous agents that defines ethical admissibility without optimization, utility functions, or probabilistic reasoning. The central object is Ω, the region of admissible actions, defined as the intersection of four structural constraints derived from four axioms. An agent with LETOS does not choose the best action — it verifies membership in Ω. The theory proves T1–T8 and P-21 through P-26, including complete boundary characterization, structural separation of authority and belief, and multi-agent coordination without knowledge merge.

Ask AI

Helpful

Bookmark

View Full Paper

Ask AI

Helpful

Bookmark

View Full Paper

Epistemic Safety Without Optimization: A Structural Theory of Ethical Admissibility for Autonomous Agents

Key Points

Abstract

Cite This Study

Also Consider

Also Consider