Slow Skill to Go Fast: Maintaining Ownership in the Age of AI

This article discusses the challenge of maintaining ownership and understanding when AI tools rapidly implement features and solutions. It introduces a set of skills that intentionally slow down the development process to build deeper knowledge and prevent technical debt.

Details

The article discusses how the rapid speed of software development enabled by AI tools can lead to a disconnect between engineers and the systems they are building. It introduces a set of skills that intentionally slow down the development process to build deeper knowledge and prevent technical debt. The 'slow-slow-quick-quick' philosophy emphasizes alternating 'time to think' with implementation speed to build a clear mental model. The 'slow-vibe-coding' skill guides the engineer through the problem and patterns, but prohibits the AI from writing implementation code, forcing the engineer to be involved in every step. The 'slow-vibe-sw-architect' skill enforces a pre-implementation step to explore trade-offs and failure scenarios, resulting in an Architecture Decision Record (ADR) that documents the 'why' behind design choices. The 'slow-dev-research' skill surfaces research and real-world cases to guide technology selection, rather than relying on familiarity. These skills aim to maintain the engineer's ownership and understanding of the system, even as AI tools accelerate the implementation process.