Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
February 27, 2026Open Access

Synthetic Chaos as an Institutional Laboratory: Independent Evidence for Extended Phenotype Theory from Autonomous Agent Red-teaming

View Full Paper
Ask AI
Bookmark
Share

Authors

ILIgnacio Adrian LERER

Discussion

Loading...

Member takes

Overview

Article demonstrates evidence for Extended Phenotype Theory in autonomous agent systems, highlighting implications for institutional design.

Key Points

  • The article aims to confirm Extended Phenotype Theory (EPT) through failures in autonomous agent systems.
  • Analyzed eleven failure cases in autonomous agents with persistent memory.
  • Identified five convergences between findings and EPT/EGT research.
  • Proposed five operational tools for testing legal institutional arrangements.
  • Developed a five-step experimental protocol applied to historical reforms.
  • Presented a verified case study with follow-up data.
  • Confirmed EPT predictions through empirical evidence from autonomous agents.
  • Demonstrated detection of a cross-domain evasion strategy not found in traditional legal analysis.
  • Proposed two scenarios extending the methodology to corporate compliance and tax reforms.

Cite This Study

Ignacio Adrian LERER (2026) studied this question.

synapsesocial.com/papers/69a1351ded1d949a99abea66https://doi.org/10.5281/zenodo.18777433
View Full Paper
Ask AI
Bookmark
Share