THE INFRASTRUCTURE OF AN EMERGENT SYNTHETIC SOCIETY
Three connected papers examine the foundations of synthetic society, joined by independent ESI reports that investigate consequential claims at the frontier of artificial intelligence.
RESEARCH PAPERS / SERIES 001 + SPECIAL REPORTS
Read the connected series in sequence or open a special report on a current frontier claim. Each entry includes a responsive article or paper reader and a direct link to its original PDF.
A fact-check and analysis of GPT-6 Astra, its extraordinary benchmark results, and the evidence for—and against—calling the current frontier of AI an AGI milestone.
This paper explores the knowledge graph as the epistemic control plane for a synthetic society: a shared external memory, semantic address space, provenance system, temporal world model, trust fabric, and coordination substrate. Through this layer, AI agents can determine what they know, why they believe it, whom they trust, what has changed, and which actions they are permitted to take.
This paper presents an architectural exploration of distributed cognition and shared memory for collaborating AGI agents. It emphasizes hierarchical memory, modularity, inter-agent communication, memory consistency, hybrid plaintext, activation, and parametric memory, lifecycle management, and continuous adaptation at scale.
This paper explores an AI-agent knowledge graph functioning as the cognitive nervous system of an AGI society. It enables decentralized knowledge exchange, layered working and verifiable memory, provenance, meta-reasoning, cross-agent auditing, adaptive collaboration, and the emergence of collective intelligence.