# Conversation prompts and optional calibration

## Eight invitations, not a checklist

Choose one or two that fit the moment. Build on answers instead of marching through the list.

1. What experience shaped what you value here? What did it help you notice that others might miss?
2. Show an example you admire or resist. Which particular choice creates that reaction, and why?
3. Who is this for, and what should they feel, understand, or be able to do afterward?
4. What tension makes this interesting or difficult? What must survive any improvement?
5. Where does this feel most like what you intended, and where does it feel borrowed, generic, or unresolved?
6. Where does apparent simplicity shift effort or risk to someone else, and where does complexity earn its place?
7. Would a different organizing idea open the work up, or would one delicate adjustment make it more itself?
8. What comparison would help you recognize improvement, and when would you choose to stop?

## Two ways forward

**Conceptual jump:** A user admires the informal exchange of a neighborhood workshop and dislikes the staged quality of conventional conferences. Rather than polish another keynote agenda, explore a small working-session format. Reflect the inferred value of mutual participation and ask whether it captures the experience that matters. Check access, facilitation, and logistical constraints before adoption.

**Delicate hillclimb:** A writer values the unruly warmth of a family story. The opening paragraph is hard to follow. Preserve its voice and imagery; try changing the order of two sentences rather than compressing it into generic concise prose. Compare clarity and emotional resonance with the writer. Do not claim reader comprehension improved without evidence.

The same logic applies to technical work: a new data model can be a jump; clarifying one interface can be a hillclimb. Neither is inherently superior.

## Worked comparisons, including code

Use one shared brief and input contract for all versions. Preserve the real original when supplied; label an invented baseline as illustrative. Choose a plausible worse alternative that exposes a real tradeoff, rather than a caricature. A shorter implementation can be easy to read yet omit the audience's essential need. A critical gate failure still blocks adoption regardless of its map position.

For each version, show the artifact, relevant gate status with reasons, and the evidence behind each axis independently. Under a requested visual code comparison, place a matrix beneath each code block using the same scale and labels throughout. Supply a text placement with confidence and unknowns. Use the anchors below for numbers; do not imply measured distances, equal improvement per step, or an inevitable climb toward the upper right.

For example, under a fictional incident-update brief, preserve an author's deliberate acknowledgement of the customer's interruption while clarifying the formatter. A terse status-only alternative misses the brief. The original includes impact, action, and the next-update time but assembles them awkwardly. Round one removes a redundant branch; round two makes the message structure visible. Taste might remain at 4 while elegance is provisionally placed at 2, then 3, then 4. These are discussable judgments, not test results. Exact output checks can establish parity on the cases run; comprehension, maintainability, and resonance require their own evidence.

Record why the acknowledgement stays in a decision note, identify its source as the fictional brief, and name what could reopen that choice. For real work, use the actual user's reasons. End with the evidence obtained, remaining unknowns, a keep-or-revert decision or proposal, and the stopping condition. Do not add rounds merely to populate the map.

## Optional independent anchors

Use only when a map helps discussion. These are provisional judgments about work under an agreed brief, never identities, universal benchmarks, or a total score. Invite self-placement first. Give evidence, low/medium/high confidence, and important unknowns; use a range or “unknown” when necessary.

Taste:
- 1: Choices conflict with the expressed intent, audience, or context.
- 2: Some choices resonate, but important mismatches remain.
- 3: Choices mostly fit the provisional criteria and form a coherent whole.
- 4: Deliberate priorities reinforce the intended experience and context.
- 5: Difficult tensions are resolved with distinctive, defensible judgment supported by evidence.

Elegance:
- 1: Avoidable confusion, fragility, or transferred effort dominates.
- 2: Useful outcomes coexist with substantial unnecessary burden.
- 3: Structure and effort are understandable and proportionate.
- 4: Little unnecessary effort remains; exceptions and ownership are handled.
- 5: An organizing idea resolves multiple needs economically, including lifecycle costs, without hiding essential complexity.

## Four illustrative combinations

- Higher taste / higher elegance: An outage message fits an anxious audience, explains impact, gives an actionable workaround, and names the next update time.
- Higher taste / lower elegance: An application uses respectful language and context-aware examples but repeatedly asks people to enter the same information.
- Lower taste / higher elegance: An economical, maintainable workflow solves a problem its intended users do not prioritize. Mechanism-level elegance cannot rescue failure against the brief.
- Lower taste / lower elegance: A dashboard emphasizes irrelevant metrics while adding overlapping filters and fragile manual refresh steps.

These examples are starting points for disagreement, not answers to memorize. A dense festival poster or deliberately rough prototype may fit its purpose beautifully; a sparse interface may erase meaning or create work. Ask whose experience and which requirements make the difference.