Practice
Five structural problems AI tools create for creative work. I built the fix for each one.
I was trained in visual communication at SVA. The whole job was reading what the audience actually receives before you design what you send. That’s attunement. I taught special education in Brooklyn after that, a self-contained classroom, twelve students, every subject. Same skill in a different room: reading what each student actually needs before the system asks them to perform. That’s accommodation. And I spent thirteen years building one enterprise platform through three technology shifts, where the hard part was never the build. It was holding the structure together while everything changed and nobody was maintaining fidelity to what had already been decided.
Those three things produced what’s on this page. Each pillar below came from one of them.
Voice Governance
Your voice is the first thing AI tools flatten.
Three drafts in and you can't hear yourself anymore. The tool smoothed your rhythm into something competent and generic. I built a system that catches the drift.
Lens Extraction
One score is not evaluation.
A single AI verdict is consensus theater. Real evaluation requires independent perspectives that can disagree. I built the framework for running them.
Structural Capture
The tool can't tell a breakthrough from a typo fix.
Your best thinking disappears when the session closes. Only you know when your mind turned. I built a protocol for marking it.
Input-First Design
The system should accommodate the human, not the other way around.
AI tools want structured input before you've finished thinking. I reversed that. You dump your thinking raw. The structure comes after.
Practitioner-in-the-Loop
The human stays the authority.
The AI industry is racing toward replacing the human. I built in the opposite direction. The tools are fast. The judgment is mine.