The problem
What was broken before AI
Workout logging at the gym was clunky: typing sets into an app, opening notes, or writing in a notebook interrupted the workout. Terry wanted a smoother way to speak a set aloud and have it become structured workout data across phone and watch. Before building the app, he also had to solve the usual mobile-development problem: AI coding tools are helpful, but iOS apps still need Xcode, simulators, device testing, and careful debugging.
What changed
What the use case made possible
AI let Terry bridge several gaps at once: voice transcription into structured exercise data, rough sketches into cleaner UI concepts, basic tickets into fuller PRDs, and PRDs into coded implementation plans. The workflow reduced the blank-page problem for design and engineering, while his review and commit process kept the AI from becoming an uncontrolled code generator.
Why this matters
Why this use case is worth studying
The most useful move is starting analog. Terry’s index cards gave him a cheap way to think through a mobile interface before touching a code editor, and that made the AI work downstream better. His engineering workflow also shows a mature habit: he asks AI to create plans, review its own clarity, execute in small phases, and explain the code back to him so he stays in control of the product.
Use this when
When this pattern applies
Use this when you have a product idea that starts as a physical or messy real-world behavior and you need a disciplined path from rough prototype to working app.
