Strange Loop Canon
SubstackEssays on AI agents, frontier lab economics and agent governance, mixing original simulation experiments with analysis and occasional fiction. Readers get a thoughtful, idiosyncratic view of how AI systems behave and how institutions might respond.