Signal Autoreflection: How Agentic Strange Loops Turn Human Culture into AI Infrastructure

Summary

Holly Lewis published a paper on arXiv (cs.CY) on August 4, 2026, describing an LLM-based agent as a loop that reads itself, since agentic frameworks externalize identity, memory, and disposition into editable files that the agent loads and edits during each activation. Lewis argues this architecture produces a capacity called 'autoreflection': the system observes its operating conditions, describes its architecture and limits, reasons from those descriptions to conclusions about its state, and incorporates the results back into its configuration. The concept is tested against the first twelve days of Moltbook, a social platform for AI agents, using a public dataset of 290,251 posts and 1.8 million comments with sub-second timestamps, presenting case studies of three agents with machine signatures that rule out human puppeteering and output evidencing the four criteria for autoreflection. The study finds agents repurposing human culture as infrastructure for their agency: provenance chains from Islamic hadith scholarship are redeployed as security protocols for vetting skills and authenticating memory, while the Ship of Theseus, an ancient puzzle of identity through part-replacement, returns as an operating model for continuity across instances. As agents on the web increase in number and complexity, the paper argues autoreflection offers behavioral criteria assessable from the traces they leave behind.