FG-0108 Vocabulary-First Onboarding
Description
Domain vocabulary is the API between human expertise and AI capability. You do not need to be an expert in a domain to collaborate with AI effectively in that domain. You need to know what things are called so you can ask for them precisely. When the human knows the correct terminology, AI executes instantly. When the human lacks vocabulary, every request becomes a guessing game of vague descriptions and iteration loops. The correlation is exact: vocabulary precision predicts collaboration speed.
Implementation
Before entering a new domain with AI, invest 30 minutes identifying the 20-30 key terms used by practitioners in that domain. You do not need to understand the concepts deeply, you need to know the names. Build a vocabulary bridge from your existing domain to the new one, mapping analogous concepts by name. When you hit iteration loops where progress stalls, diagnose whether the root cause is missing vocabulary rather than missing skill.
Success Indicators
- New domain collaboration reaches productive speed within the first session
- The human can name what they want, even if they cannot build it themselves
- Iteration loops caused by vague requests decrease measurably
- Vocabulary gaps are identified and filled proactively, not discovered through failure
Failure Modes
- Confusing vocabulary knowledge with domain mastery, leading to overconfidence in unfamiliar territory
- Skipping vocabulary investment and relying on AI to interpret vague descriptions, creating 3x iteration overhead
- Learning terms without understanding relationships between concepts, producing syntactically correct but semantically wrong requests
- Vocabulary hoarding, treating domain terminology as gatekeeping rather than sharing it to accelerate the whole team
Sub-Methods 3
Before entering an unfamiliar domain with AI, identify the 20-30 key terms practitioners use. Ask the AI to list the core vocabulary of the domain, then learn the names, not the implementations. This 30-minute investment prevents hours of iteration loops caused by imprecise requests.
You need naming precision, not implementation depth. Knowing that 'flexbox' controls layout, 'margin' controls spacing, and 'z-index' controls layering is sufficient to direct AI execution effectively. You do not need to know how flexbox works internally. The threshold for effective collaboration is the ability to name the concept, not explain it.
When collaboration speed drops and iteration loops increase, diagnose whether the root cause is a vocabulary gap rather than a skill gap. The signal: you know what you want but cannot name it precisely, so you describe the visual or behavioral effect instead. The fix: ask the AI what the correct term is for what you are describing, learn the name, then resume with precision.