BACKGROUND
This project was a human-in-the-loop example demonstrating a specific AI use pattern: using AI to consolidate research and build a structural scaffold for a thought leadership piece, then writing the piece entirely from scratch using that scaffold as context. The goal was to show clearly where AI contributed value and where it would have reduced it.
I documented the full process across three stages: the AI research summary and structural outline, the human-written article, and a comparison table mapping specific elements of the published piece to what the AI scaffold provided versus what the human rewrite added and why the distinction mattered. The table included the opening observation, the mechanism of failure, the recommendation section, and the acknowledgment of difficulty, showing in each case how the AI version was accurate but abstract while the human version was grounded in direct practice.
The result was a document that made a specific argument: the scaffold saved two to three hours of research and structuring time, but the piece's value, the reason a senior practitioner would share it with their team, was entirely in the writing, and submitting the AI scaffold as the piece would have been technically adequate and completely forgettable.
