2/20/2025
Published on Convert on 20/02/2025.
Letting AI “do it all” in one prompt often feels generic and out of your control. This guide (Part 1) gives frameworks so AI amplifies your work—used in workshops and in tools like Ressada and Experimentation.wtf.
Scaffolding: Map the creative process as Explore → Concept → Delivery. Add milestones (e.g. idea, research, outline, draft, edit) as goals for each AI step.
How LLMs work (simplified): They pattern-match. You can think of the prompt as one “panel” and the output as the next; where you put attention in the prompt shapes what they do. You stay in charge by choosing what to emphasise.
Three modes:
- Human-centric: You lead; AI validates, fills gaps, or checks bias. Best for final delivery.
- AI-centric: AI leads; you use the output (e.g. summarise, brainstorm). Best for early explore/concept; risky for final delivery.
- Synergistic: Back-and-forth, shared decisions. Powerful but needs more structure.
Kolb cycle: After each AI interaction—Experience → Reflect → Conceptualise → Experiment, then repeat. That loop keeps you in the driver’s seat and gives quick feedback. Part 2 adds concrete Interaction Patterns (beats and sequences) for research, ideation, and critique.
