Definition: Building software or tools with AI the way a scribe writes, with intent, craft and a clear record of what you meant. The person stays in charge by describing clearly what they want and why, keeping authorship and understanding at the centre rather than letting the AI carry the whole task.
So what? It sits as a counterpoint to vibe coding, where output tends to be accepted on feel. Prioritising clear intent and a deliberate record of decisions makes the resulting work easier to test and maintain. It's leans towards quality-in-the-loop instead of human-in-the-loop.
Examples: Capturing a team’s house voice, process and intent as a reusable AI skill, so the output reflects what was actually meant rather than a plausible guess; building a dashboard through clear and deliberate direction so each design and data decision is understood and recorded rather than guessed.
So what? It sits as a counterpoint to vibe coding, where output tends to be accepted on feel. Prioritising clear intent and a deliberate record of decisions makes the resulting work easier to test and maintain. It's leans towards quality-in-the-loop instead of human-in-the-loop.
Examples: Capturing a team’s house voice, process and intent as a reusable AI skill, so the output reflects what was actually meant rather than a plausible guess; building a dashboard through clear and deliberate direction so each design and data decision is understood and recorded rather than guessed.
