SlopNet: The Digital Scorched Earth Strategy — A Theory on Convenient Data Contamination

This essay presents "SlopNet" as a theoretical framework developed by IA-ismo LAB to analyze a series of events that, at minimum, appear remarkably convenient for Big Tech: the systematic contamination of the open web with AI-generated synthetic data (AI Slop). We do not claim this is a deliberate strategy. We observe that model collapse and data poisoning—whether intentional or not—function as a competitive moat that disproportionately benefits first-generation model owners with clean pre-2023 training data. The convenience of these outcomes for incumbent players is, at the very least, worth examining. The analysis explores three fronts: (1) the initial web harvest period (2020-2023), (2) current extraction operations through protocols like MCP targeting professional workflows, and (3) the documented plundering of cultural heritage. Case studies include the OpenClaw/$CLAWD token incident and Anthropic's internal documents revealed by The Washington Post. This is a speculative-critical exercise—not a peer-reviewed paper—intended to provoke discussion about structural incentives and convenient coincidences in AI development.