Dark Factories: Rise of the Trycycle

2026-04-21 · Source: AI & ML – Radar · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning · Depth: Intermediate, medium

Summary

The article introduces the concept of "dark factories," AI-powered engines that automate software production from specifications. These factories leverage the principle that AI performance improves with increased usage, often employing two patterns: "slot machine development," which involves querying multiple AIs simultaneously for the best output, and the "trycycle" pattern, a simple iterative loop of planning, implementing, and refining. Several implementations are highlighted, including Steve Yegge's "Gas Town," a complex MMORPG-like system for code generation; StrongDM's "Attractor" specification, which defines a feedback loop where models improve by processing their own output; and "Kilroy," a Go-based implementation of the Attractor. The article also presents "Trycycle," a minimalist skill for agents like Claude Code and Codex CLI, designed for rapid feature delivery by iteratively defining, planning, and implementing solutions.

Key takeaway

For Machine Learning Engineers seeking to accelerate software delivery, consider integrating a "dark factory" approach. If you need immediate results, deploy the Trycycle skill with your coding agent for rapid feature implementation. For more complex, configurable systems, explore Kilroy or the StrongDM Attractor specification to build a tailored, continuously improving development engine.

Key insights

AI-driven "dark factories" automate software development through iterative refinement and parallel AI querying.

Principles

• AI improves with more usage.
• Parallel AI queries yield better results.
• Iterative refinement drives software development.

Method

The "trycycle" method involves defining a problem, writing a plan, iteratively refining the plan, implementing it, and then iteratively refining the implementation until perfect.

In practice

• Use Trycycle skill for rapid feature delivery.
• Implement StrongDM Attractor for custom factories.
• Explore Gas Town for collaborative code generation.

Topics

• Dark Factories
• AI Software Development
• Trycycle Pattern
• StrongDM Attractor
• Kilroy

Code references

• danshapiro/kilroy
• obra/superpowers
• danshapiro/trycycle

Best for: Machine Learning Engineer, AI Engineer, Software Engineer, Director of AI/ML

Related on AIssential

• An AI state of the union: We’ve passed the inflection point, dark factories are coming, and automation timelines | Simon Willison — AI has fundamentally reshaped software engineering, making code generation cheap and amplifying experienced engineers' capabilities.
• The loop of progress — Progress stems from relentless, iterative refinement through ideation, experimentation, and analysis.
• The Production AI Playbook: Deploying Agents at Enterprise Scale — Sandipan Bhaumik, Databricks — A structured five-pillar framework is essential for successfully deploying and managing enterprise-scale AI agents in production.
• Building with AI is easy. Shipping is hard — Building AI prototypes is easy; achieving reliable, efficient, and modular production systems requires a specific methodology.
• Vibe Graphing: 10x More Affordable than Vibe Coding (MAS-Factory) — Vibe Graphing automates multi-agent system design from natural language, drastically cutting costs and development complexity.
• Everyone Is Chasing AI Agents. I’m Learning to Build Boring Systems — Real-world AI systems demand robust "boring" infrastructure like validation and logging, not just impressive agent capabilities.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI & ML – Radar.