Navigation

2. Create

Build the smallest working artefact capable of generating real behavioural evidence. Not decks — something real.

Create converts a framed hypothesis into an executable artefact. Once the problem is sharply defined, Create makes it tangible. This mode combines human design judgement with AI-assisted generation to build the smallest artefact capable of producing meaningful feedback.

In the AI era, this may include interactive prototypes, thin-slice coded MVPs, workflow simulations, or service experiments. Design and code do not operate sequentially — they co-exist. The goal is not polish or completeness, but testability. If an idea cannot be expressed in executable form, it cannot be validated.

Speed of creation is no longer the constraint. What you choose to create is. The discipline is building the minimum version that will actually teach you something.

Activities

  1. Parallel build sprint — Generate 3–5 fast directional concepts using human judgement and AI generation. Select by testability, not polish. The question is not "which is best?" but "which will teach us the most?"
  2. Working build — Translate the selected direction into an interactive prototype or thin-slice build. AI tools generate copy, interface, and data scaffolds.
  3. Experience refinement — Test flows internally (heuristic or agent-based tests). Visual polish to make it feel effortless. Ensure technical feasibility.
  4. Concept review — Stakeholder playback: "Does this feel like the future?" Output: high-fidelity prototype + experience rationale.

Outputs

  • Interactive prototype or thin-slice working build
  • Clear mapping from hypothesis → solution components
  • Documented assumptions embedded in the artefact
  • AI-assisted technical and design documentation
  • Defined validation plan (what will be measured and how)

The objective is not completeness. The objective is testability.

Signal: Test Readiness

You can move on to Validate once:

  • The artefact can generate behavioural evidence (not just opinion).
  • Core assumptions are embedded in something executable.
  • Measurement mechanisms are defined (telemetry, interviews, KPIs).
  • The team agrees the artefact is sufficient to expose to reality.

If the artefact cannot generate a measurable signal, it is not ready.

3. Validate→