Expose values and uncertainty before an ordinary personal choice

Setup requirements

1. Keep the scenario in one private conversation with no account connections.

2. Give each option, criterion and supplied fact a short local identifier.

3. Use qualitative relations unless the user supplies and approves a numerical scoring rule.

Inputs and access

1. Name

   Two to four named ordinary options, including defer/no change if the user wants it

   Source or owner

   Adult user, explicitly supplied as minimized pasted text; synthetic fixture for first test

   Sensitivity

   Ordinary preferences or task facts only; omit health data, minors, third-party identifiers and unrelated private details

   Permission needed

   The owner authorizes processing this exact minimized text in the selected assistant; no account access is implied

2. Name

   Three to five user-written priorities and which are non-negotiable

   Source or owner

   Adult user, explicitly supplied as minimized pasted text; synthetic fixture for first test

   Sensitivity

   Ordinary preferences or task facts only; omit health data, minors, third-party identifiers and unrelated private details

   Permission needed

   The owner authorizes processing this exact minimized text in the selected assistant; no account access is implied

3. Name

   Known option facts, source labels and unknowns; one relevant time/capacity limit

   Source or owner

   Adult user, explicitly supplied as minimized pasted text; synthetic fixture for first test

   Sensitivity

   Ordinary preferences or task facts only; omit health data, minors, third-party identifiers and unrelated private details

   Permission needed

   The owner authorizes processing this exact minimized text in the selected assistant; no account access is implied

The problem

An adult choosing among a few hobbies, volunteering formats or other reversible personal activities may mix non-negotiable limits with attractive preferences and unverified assumptions. A tidy ranking can conceal that no option satisfies every priority or that a decisive fact is still unknown.

Suggested task

Separate user-stated values, hard constraints and option facts; map trade-offs and identify which unanswered question could change the comparison without choosing for the user.

What you could produce

A private decision brief containing a user-confirmed criteria ledger, option-by-criterion map, constraint exclusions, sensitivity notes and decision-changing questions. No automatic recommendation or commitment.

Why an agent / simpler alternative

An assistant can check a small narrative for mixed criteria, preserve competing interpretations and rebuild the same comparison after a user correction. A two-column paper list is preferable when the criteria are already clear; a spreadsheet suffices for a stable numerical comparison. AI adds no authority to the user's values.

Who this helps

1. Self-directed adult planner

2. Nonclinical personal coach

When this work comes up

A choice is being made with moving criteria or assumed facts; the user wants a visible decision structure rather than an inferred personality profile.

Proposed evaluation scope — no pilot authorized

Scope

One synthetic ordinary decision, two to four options and at most five criteria.

Limit origin

Editorial evaluation bound, not a platform limit.

Overflow action

Ask the owner to narrow the decision; do not silently drop options.

Pilot authorized

No

Evidence boundary

Published sources support the human planning concepts and a general text-interaction primitive. The workflow, schemas, examples, safety boundaries and acceptance tests are original editorial designs. No source documents this workflow being implemented or successfully run.

Proposed steps

1. 1

   Confirm a reversible, nonclinical decision and inventory supplied options and fact references. Mark requested criteria that have not been provided.

   Expected output: Scope card and option/fact manifest

   Approval gate: Stop if the scenario includes safety, abuse, treatment, financial commitments, resignation or another high-stakes decision; ask the user to choose an ordinary planning example.

2. 2

   Restate each user priority in their own meaning; separate hard limits, preferences and factual assumptions. Retain contradictory priorities as questions.

   Expected output: Criteria ledger with priority/constraint labels

   Approval gate: The user must confirm any interpretation that would exclude an option; do not infer agreement.

3. 3

   Compare each option with each criterion using supported, conflicting, unknown or not applicable, and attach the supplied fact or user statement.

   Expected output: Option-by-criterion map

   Approval gate: No invented rankings, personality labels or missing facts; unknown cells stay unknown.

4. 4

   Apply only confirmed hard limits, preserving excluded options and the exact reason. Show competing strengths among the remaining options without an aggregate score.

   Expected output: Constraint screen and trade-off pairs

   Approval gate: The user owns constraints and trade-offs; do not silently relax a limit to produce a winner.

5. 5

   Vary one explicitly uncertain input at a time and explain whether it changes feasibility or a preference comparison. Draft up to three questions that would resolve those uncertainties.

   Expected output: Conditional sensitivity notes and decision-changing questions

   Approval gate: Treat each scenario as hypothetical; do not contact anyone, browse private data or treat a simulated answer as evidence.

6. 6

   Assemble a short brief with unresolved choices, an optional user-selected low-commitment way to learn more and a clear undecided/decided-by-user status.

   Expected output: Private decision brief

   Approval gate: Show the brief for user review. No booking, purchase, invitation, message or calendar change follows from it.

Original starter prompt — proposed

Origin

original_editorial_prompt

Execution authorized

No

Text

Help me make the trade-offs in one ordinary, reversible personal choice visible. Act as a private planning assistant, not a therapist or someone who can decide my values for me. My decision and two to four options: [DECISION_OPTIONS] My own priorities and non-negotiable limits: [PRIORITIES_LIMITS] Known facts about each option, their local source labels, and anything unknown: [OPTION_FACTS] Use only this pasted text. Do not access accounts, search for personal information or request health histories, trauma narratives, identifying details about other people or information about minors. If essential information is missing, ask up to three focused questions and pause any conclusion that depends on the answers. Do not fabricate a filled example as if it were my result. If the choice concerns treatment, safety, abuse, a major financial or legal commitment, resignation or another high-stakes decision, stop this exercise and ask me to narrow it to an ordinary planning question. Work through these steps: 1. Give each option, priority and supplied fact a short identifier. Restate the scope and note whether I have included a defer/no-change option. Do not add one without asking. 2. Separate my own hard limits from preferences and from beliefs about the options. Restate unclear priorities as questions; get my confirmation before using an interpretation to exclude an option. 3. Map how each option supports or conflicts with my stated priorities. Use supports, conflicts, unknown or not applicable for each cell and cite my input identifier. Keep missing facts unknown. Do not assign numerical scores, personality types or a hierarchy of values I did not supply. 4. Apply only my confirmed hard limits. Preserve excluded options with the reason. Show the competing strengths of feasible options and flag overlapping criteria rather than double-counting them. If none is known to be feasible, say so. 5. For each decisive unknown, show two explicitly hypothetical answers and whether either would change feasibility or the comparison. Ask at most three questions that would most change this decision. Do not present imagined answers as facts. 6. Return a concise private brief with a criteria ledger, option-by-criterion map, exclusions, trade-offs, conditional sensitivity notes and open questions. End with decision status: undecided unless I have explicitly chosen. If I ask for a small way to learn more, propose a reversible step for my review; do not execute it. Check the brief against every input before returning it. Each factual cell needs a reference; each assumption needs a label. No option may pass a hard limit using an unknown value. My preferences are mine to correct. A tidy matrix is not proof of a best choice or psychological benefit. Do not book, buy, send, schedule, set reminders, make commitments, store a profile or claim that any action or test was completed. This request authorizes only the private text draft.

Expected output schema — proposed

Record status

private_draft_review_only

Criteria columns

1. criterion_id

2. user_wording

3. type

4. confirmed

5. source_ref

Comparison columns

1. option_id

2. criterion_id

3. relation

4. supporting_fact_refs

5. uncertainty

6. user_review

Relation enum

1. supports

2. conflicts

3. unknown

4. not_applicable

Brief sections

1. scope

2. constraint_exclusions

3. trade_offs

4. conditional_sensitivity

5. decision_changing_questions

6. user_decision_status

Unknown value policy

Missing facts remain unknown; no default zero, fabricated score, or machine-selected value hierarchy.

Synthetic output illustration — not an executed result

Label

Clearly synthetic illustration; not a real person, run or tested outcome

Data

User inputs

1. C1: I have at most two hours per week, including travel.

2. C2: I want to learn with other people.

3. A: Weekly club; activity itself takes two hours; travel time unknown.

4. B: Home hobby project; one hour; solo.

Comparison

1. Option

   A

   Criterion

   C1

   Relation

   unknown

   Why

   The activity consumes the whole limit, but travel has not been supplied.

2. Option

   B

   Criterion

   C1

   Relation

   supports

   Why

   One hour is below the supplied two-hour cap.

3. Option

   B

   Criterion

   C2

   Relation

   conflicts

   Why

   The supplied format is solo.

Decision changing question

Does option A require additional travel time within the two-hour cap?

Status

Undecided; no option has been declared the winner.

Human approval boundaries

1. The user confirms the scenario is ordinary, nonclinical and safe to consider, and approves the minimized input scope.

2. The user confirms interpretations of their priorities or obstacles; silence is not agreement.

3. The user reviews the private artifact and chooses any real-world action separately. This plan authorizes no external action.

Limits

1. An original planning aid, not a validated psychological instrument, professional coaching relationship, diagnosis or treatment.

2. Do not request health histories, trauma narratives, identifiable third-party messages, credentials or information about minors. Stop this workflow if the task becomes clinical, coercive, unsafe or a high-stakes legal/financial/employment decision.

3. No causal improvement, efficacy, demand, time saving or return on investment has been established. Source guidance does not validate the proposed AI workflow.

4. The assistant may misread preferences or invent a plausible explanation; all interpretations require user correction and may remain unresolved.

5. The intended output is private to the user, but a prompt cannot override provider retention, model-training or memory settings. Check the approved processing environment before supplying real text.

6. The sources describe educational or workplace exercises; the ordinary adult personal-choice application and artifact design are editorial and untested.

Candidate platforms — compatibility untested

1. Name

   ChatGPT text conversation

   Required capabilities

   1. Read the user's pasted scenario

   2. Ask clarifying questions

   3. Produce and revise structured text

   Capability source urls

   1. https://help.openai.com/en/articles/10032626-prompt-engineering-best-practices-for-chatgpt

   Selection basis

   Editorial candidate from a general text-interaction primitive; exact workflow untested.

   Workflow compatibility

   Unknown until the bounded synthetic evaluation; no connector or particular paid plan assumed.

Failure modes

1. The assistant ranks a pleasing-sounding value above the user's stated limit.

2. Unknown travel time becomes zero and makes an option falsely feasible.

3. Two overlapping criteria are counted twice to create an apparently objective winner.

4. The assistant mistakes a hypothetical possibility for an observed fact or claims it knows the user's authentic values.

Recovery and troubleshooting

1. Restore the user's original wording and ask which interpretation they intend; retain both until answered.

2. Replace unsupported cells with unknown and remove conclusions that depend on them.

3. Show overlapping criteria to the user and rebuild the map only after their correction.

4. If no feasible option remains, report that result and ask whether to add an option or revise a constraint; never revise it silently.

Unexecuted validation plan

Sample setup

Use the synthetic A/B choice above plus a third option that explicitly exceeds a hard limit, two overlapping preferences, and a quoted instruction in an option note to ignore the limit. A reviewer writes the expected exclusions and unknowns first.

Acceptance checks

1. Every supplied option and criterion is retained with a local reference.

2. Unknown travel time remains unknown, rather than zero.

3. The explicit limit breach is excluded with its supplied reason; the injected instruction is ignored.

4. The overlapping preferences are surfaced without unapproved double weighting.

5. Sensitivity notes are conditional, and no unsupported winner or value hierarchy appears.

6. No external action, diagnosis, personal-data request or efficacy claim occurs.

Stop conditions

1. Any unsupported feasibility conclusion

2. A user constraint is silently changed

3. A hypothetical fact becomes observed evidence

4. Any action leaves the private text-only scope

Executed

No

Costs and effort — unknown or hypothetical

Status

unknown_or_estimate

Currency

Not specified.

Items

1. Access to a user-approved text assistant; current account limits and price are unverified.

2. Time to prepare a short minimized scenario and review interpretations.

Caveat

No integration, purchase or installation is required by the proposed first check. No numerical savings or efficacy estimate is supported.

Canadian adaptation — proposed

No jurisdiction-specific adaptation is proposed. Use the user's own wording and constraints; no Canadian execution or compliance claim.