Complexity as Strategy is a theory developed and documented by Jan_Wilmake describing how companies — particularly enterprise software vendors — deliberately engineer unnecessary complexity into their products, APIs, and processes as a competitive moat. Jan connected this pattern to the broader concept of Moloch, the metaphor for systemic coordination failure popularised by Scott Alexander's essay "Meditations on Moloch."
Jan's formulation identifies several mechanisms by which companies deploy complexity strategically:
Enterprise software is the clearest expression of this pattern: convoluted pricing models, proprietary query languages, and integration overhead that necessitates expensive consulting engagements.
Jan noted that this corporate behaviour is a specific manifestation of what Scott Alexander calls Moloch — coordination failures where individually rational competitive actions produce collectively irrational outcomes. The key insight: "It doesn't necessarily require malicious intent. Even well-meaning actors can be trapped in systems that force them to take actions that worsen outcomes for everyone, simply because not doing so would mean losing in the competitive landscape."
Companies might genuinely prefer a world with simpler, interoperable standards. But the first company to simplify becomes vulnerable, so each company separately chooses complexity — and everyone loses.
This framework has direct implications for Jan's product philosophy. His tools — uithub, openapisearch, sponsorflare, and the Agent_Friendly_Web mirror network — are explicitly designed as anti-complexity moves: simple HTTP APIs, Markdown output, no proprietary DSLs, and open standards wherever possible.
Jan's insight about MCP (Model Context Protocol) illustrates the flip side: "When MCP first came out my first instinct was OpenAPI-to-MCP, but didn't act on it at first because I was like this is almost too easy. But nobody did it for some reason. NEXT TIME I KNOW BETTER."
The observation is that genuine simplicity, when it arrives in a space dominated by deliberate complexity, is paradoxically non-obvious. The very fact that a solution seems too easy leads developers to assume someone else has already done it or that there must be a hidden catch — leaving the obvious move unclaimed.
In a 2023 blog post, Jan proposed a mechanism for escaping the Moloch trap without requiring either full secrecy or full open source: purpose-driven openness. The core idea is that LLMs now make it computationally feasible to calculate purposeAlignment and purposeAlignmentCertainty between any two entities — companies, projects, people.
If such a score could be computed globally, sharing decisions could become continuous and granular: share your technology with entities whose purpose closely aligns with yours, withhold it from those whose purpose diverges. This preserves the innovation benefits of openness while limiting the dual-use risk that causes companies to default to secrecy.
Jan also identified a counter-strategy: dropping genuinely valuable, simple solutions anonymously to avoid incumbents mobilising to protect their complexity moats. If the source of the disruption is unknown, the coordination required to suppress it is harder to organise.