10-10-10

Activate when: someone is making a decision under emotional pressure and worries they'll regret it later; someone says "I know I should do X but it feels too scary right now"; someone is about to quit, accept an offer, end a relationship, or have a hard conversation and is stuck; someone mentions "regret minimization," "what would future me say," or "Suzy Welch." Do NOT activate when: the decision has no medium- or long-term consequences (e.g., which lunch to order); the decision belongs to someone else and the user is being asked to ratify it.

Overview

Faced with a non-trivial decision, ask three questions: How will I feel in 10 minutes? In 10 months? In 10 years? The three horizons are spaced an order of magnitude apart to surface the systematic bias toward the immediate — hyperbolic discounting — that makes decisions under stress go wrong. Coined by Suzy Welch (2009); the 10-year horizon is operationally identical to Jeff Bezos's regret-minimization framework (1994).

Composes with regret-minimization, second-order-thinking, first-principles, and metacognition.

When to Use

• A personal life decision is being made under emotional pressure
• A career or professional choice has long-term consequences that current anxiety is obscuring
• A relationship decision (start, deepen, end) requires weighing what feels acute vs what will matter
• A high-stakes business call is being made by someone who would benefit from a forced retrospect
• An impulsive action is about to be taken that would feel embarrassing later
• A hard conversation is being deferred because it's uncomfortable now
• Someone says "10-10-10," "regret minimization," "what would future me say," "Suzy Welch," "Bezos framework," "long-term perspective"

Not when: the decision genuinely has no medium- or long-term consequences (which lunch to order); the 10-10-10 frame produces three identical answers (low-stakes — stop deliberating); urgent operational decisions where deliberation cost exceeds the framework's value; decisions someone else has made and you're being asked to ratify.

Coaching Novices (Adaptive Front Door)

• Engine mode: user has a concrete case → run The Process directly.
• Coach mode: user is unfamiliar or has no concrete case → guide step by step.

In Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.

1. One-line what-it-is: imagine yourself in 10 minutes, 10 months, and 10 years — notice which version of you would regret which choice.
2. Check fit against When to Use / When NOT to use.
3. Elicit the specific decision.

[WAIT — do not advance until user responds]

4. One question at a time: how will you feel in 10 minutes? 10 months? 10 years? Which horizon are you currently privileging?

[WAIT — do not advance until user responds]

5. Close: name the specific action the neglected horizon recommends.

[WAIT — do not advance until user responds]

The Process

S1 — State it: Decision (specific) · Options · Time pressure · Emotional state going in.

S2 — 10 Minutes: How will I feel? What emotion is driving this answer?

S3 — 10 Months: How will I feel? What will have changed? What second-order effects begin?

S4 — 10 Years: Defining moment or trivial in retrospect? What does my 80-year-old self say?

S5 — Neglected horizon: Which horizon was I privileging? Which was I neglecting? What does the neglected one say?

S6 — Commit: Action · Timeline · Early signal I made the right call.

Output Template

Decision | Options A/B | Emotional state going in | 10 Minutes (feeling + emotion) | 10 Months (feeling + what unfolds) | 10 Years (feeling + defining/trivial) | Neglected horizon + what it says | Decision: action / timeline / early signal

→ Method in Action: Suzy Welch, 1990s; Jeff Bezos, 1994

Pack: 10-10-10 Patterns

Decision type	10-min dominates when	10-year dominates when
Career change	Loss-aversion of current salary	Trajectory of capability + meaning
Relationship conflict	Anger / desire to "win"	Trust and partnership over time
Hard conversation	Discomfort of having it	Cost of letting issue compound
Health / fitness	Immediate convenience	Capability and longevity
Major purchase	Excitement of acquisition	Total cost of ownership
Saying yes to obligation	Politeness / social pressure	Time and energy reservoir

Applying It Well

• The medium horizon (10 months) is often the most informative — don't skip it.
• If all three horizons return "fine," stop deliberating; the decision is genuinely low-stakes.
• The framework ends in action — state the decision, the timeline, and the early signal.
• If you genuinely can't imagine the 10-year view, gather more structural information before deciding.

→ Primary sources: references/sources.md

Common Rationalizations

[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.

Fake move	Reality
[D] Apply 10-10-10, then ignore the 10-year answer	The most common failure. The 10-year horizon is what the framework is for.
[D] Treat 10-10-10 as if it produces an answer automatically	The framework produces information. Synthesis is still your judgment.
[D] Use 10-10-10 for trivial decisions	If three horizons return "fine," stop. This is a discipline against systematic bias, not a routine.
[D] Make up answers that conveniently align with what you wanted	Be honest. If you can't imagine the 10-year view, you may not understand the decision.
[D] Skip the medium horizon	The 10-month answer is often the most informative. Don't skip it because the other two are easier.
[D] Mistake "I can't tell" for an actual answer	If you can't tell what the 10-year horizon says, gather more information first.
[D] Use 10-10-10 to defer hard calls	The framework should end in commitment. Indefinite deferral means you were avoiding, not analyzing.
[D] Pretend 10-10-10 dissolves emotional difficulty	The framework clarifies which option is better. It does not make the better option emotionally easy.
→ Add [O] entries here after each real use — paste the actual failure pattern	What went wrong and why

Red Flags

• 10-10-10 was applied and the 10-year answer was politely ignored
• The 10-month horizon was skipped (only 10-min and 10-year considered)
• The framework was used to justify what you wanted to do anyway
• The decision was deferred indefinitely after "considering" the horizons
• The answer changes every time you apply the framework — suggesting you don't actually know your values
• All three horizons returned "fine" but you continue to deliberate

Verification

• The decision was stated specifically
• All three horizons were considered honestly
• The horizon you were privileging before was named
• The horizon you were neglecting was named
• The decision converged to specific action, timeline, and early signal
• You committed even though the better option is not the easier one

---

Part of deciqAI Knowledge Skills — open-source thinking skills that make rigor executable for AI agents. Built by deciqAI · https://deciqai.com · Contributions welcome — see the template at the repo root.