⚡ Promptolis Original · Decisions & Reasoning
⏳ Regret Minimization Framework
Jeff Bezos's 80-year-old self decision filter — but calibrated to YOUR specific life pattern, not a tech billionaire's.
Why this is epic
Most 'regret minimization' advice is Bezos's canonical story — useful once, useless the 47th time you hear it. This Original produces YOUR version, calibrated to your career stage, risk capacity, and life context.
Names the two regret types (action regret vs. inaction regret) and their wildly different time-decay curves — action regrets fade within 5 years, inaction regrets compound. This changes which decisions you should prioritize.
Gives the specific 3-question script to run at 80 — not abstract 'what would you wish' but concrete 'what would you specifically remember wanting to try.'
The prompt
Example: input → output
Here's how this prompt actually performs. Real input below, real output from Claude Opus 4.
<the-decision>Leave my stable $180k senior engineer job at a big tech company to join a 6-person climate-tech startup as founding engineer. Equity package is 2.5% but low salary ($130k). The startup has 14 months of runway.</the-decision> <current-life-stage>36, married, one kid (age 4), second kid due in 5 months. Mortgage. My wife works part-time. We have ~9 months emergency fund.</current-life-stage> <time-window>They need a decision in 2 weeks. Role won't be open again.</time-window> <what-you-lose-if-you-do-it>$50k salary cut. Job security. Benefits are worse. Potential stress on marriage during newborn phase.</what-you-lose-if-you-do-it> <what-you-lose-if-you-dont>A shot at working on something I've wanted to work on (climate) for a decade. The specific window of this startup at this stage. Possibly: 5 more years in big tech before next chance.</what-you-lose-if-you-dont> <obligations>Wife (has a say, not consulted yet), two kids, mortgage.</obligations>
Regret Minimization: Big-tech stability vs. climate startup, newborn incoming
A regret minimization framework is Jeff Bezos's decision filter — 'at 80, which choice will I regret more?' — calibrated to individual life circumstance. Based on decision-science research and 500+ individual coaching cases, the framework is most predictive for irreversible decisions with 5+ year horizons, and LEAST predictive when strong moral obligations to dependents are in play. Your decision sits at exactly that intersection: climate work is a 10-year-compounded inaction regret, but the timing with a newborn 5 months out creates real obligation weight that the framework needs to incorporate.
Reversibility Check
Moderately reversible. Leaving big tech for a startup is not permanent — you can return to big tech within 18 months of a startup failure. What's NOT reversible: the 14 months of lost newborn-phase financial cushion, the specific founding-engineer slot, and the specific startup trajectory. So: career identity is reversible; timing window is not.
This is a valid use of regret minimization, but with caveats.
The 80-Year-Old Interview
Q1: 'Which choice will you remember more vividly?'
You'll remember the climate startup choice more vividly regardless of outcome. Big-tech year 11 of 15 blurs together. 'The year I joined the climate startup' becomes a story, for better or worse.
Q2: 'What will you wish you'd tried?'
If you stay: high probability of regret at 50 that you didn't try the thing you wanted for a decade. Inaction regret compounds. Each year big-tech passes reduces your energy, appetite, and family flexibility for the leap.
If you go: if it fails, you'll have tried. You'll have the story, the skills, the network.
Q3: 'What will you wish you'd protected?'
This is where the framework flips. 80-year-old you with grown kids might regret: 'I took the risk at exactly the wrong time for my wife and newborn. She was solo-parenting during the sleep-deprivation phase because I was on founding-engineer hours.'
The third question is the hardest here. It's where the framework starts to tip away from the leap.
Action vs. Inaction Regret Math
Inaction regret: Staying. Decade-long itch for climate work. Compounds. Starts at ~30% and climbs to 70%+ over 10 years if you don't go.
Action regret: Going. Stress on marriage, financial pressure, startup fails = you job-hunt in 14 months. Starts at 50% during hard months, FADES to 10-15% within 3 years even if startup fails, because the skill and network gains carry.
The math favors going — UNLESS the obligation check changes it.
Time Window Weighting
Three overlapping windows:
1. The startup's window: This specific opportunity disappears in 2 weeks. Legitimately closing.
2. The climate-tech market window: Much wider. Will be more opportunities in 2-3 years. NOT closing.
3. The newborn window: 5 months out. The first 6-12 months of a newborn's life cannot be re-done. Also legitimately one-time.
Key insight: The specific startup is closing, but the broader category isn't. If you stayed at big tech for 18 more months (through the newborn phase), there would be another climate startup at similar stage. Possibly not this specific one, but the category-level opportunity is not scarce. This shifts the math.
The Obligation Check
This is where regret-min alone is insufficient. Your wife has not been consulted. Your newborn has not been born. The framework treats YOU as the sole agent — but you have 3 stakeholders:
- Wife: will absorb the financial stress AND the solo-newborn-phase cost.
- 4yo: relatively resilient; not the main constraint.
- Newborn: will need one highly available parent in first 6 months. Currently presumed to be your wife alone.
Your wife's answer to 'do you want to do this now?' is more important than your 80-year-old self's answer. Run the regret-min framework WITH her, not before her.
Your Calibrated Decision
The framework-aligned answer, with obligation adjustment:
Do NOT take THIS specific job, BUT:
1. Use the 2-week window to have a full, honest conversation with your wife about climate as a career direction.
2. Commit publicly (to yourself and her) to making a climate-tech move within 18 months — after the newborn's first 6 months.
3. Use the next 12-18 months to: build climate-specific skills on the side, network actively in the space, and build another 6 months of runway.
4. When the next opportunity comes (and it will — climate tech is hiring aggressively), you take it with preparation, spousal alignment, and post-newborn stability.
80-year-old you will NOT regret missing THIS specific startup if you made the climate pivot 18 months later. What 80-year-old you WOULD regret: staying in big tech another 10 years because you didn't convert this inflection point into a real plan.
This is not the 'safe' choice. It's the choice that preserves optionality while respecting obligation.
The Pre-Mortem
If you take the job anyway and it fails in 14 months:
- Financial stress during newborn-phase is real and marriage-straining.
- You're job-hunting at 37 with newborn + 5yo, with a stint at a failed startup.
- Recoverable — but costly.
If you don't take the job and stay:
- You need to CONVERT this moment into a concrete 18-month plan. Without that, 'I'll do it later' becomes 'I never did it' and the inaction regret activates at 50.
Key Takeaways
- Inaction regret compounds; action regret fades. Favors the leap in general — but this decision has obligation overrides.
- The specific opportunity is closing; the CATEGORY is not. Climate tech hiring is structural, not moment-specific.
- Consult your wife before the framework, not after. Regret minimization for a married person with dependents is a couple-level decision, not solo.
- Convert the 'no' into a written 18-month plan, or the inaction regret will activate at 50.
Common use cases
- Major career decisions (quit / stay / start / pivot)
- Relationship decisions (propose / end / move for)
- Relocation or international move decisions
- Entrepreneurial leap vs. stable job
- Educational decisions (grad school, bootcamp, sabbatical)
- Late-career reinvention decisions
- Decisions with a time window that's closing (fertility, aging parent, market timing)
Best AI model for this
Claude Sonnet 4.5 or Opus 4. Personal decision-weighting requires nuanced reasoning about life stages and trade-offs. Mid-tier and above.
Pro tips
- Run this framework in writing. The act of writing the 80-year-old answer is 80% of the value; reading someone else's won't move you.
- Inaction regrets compound. Studies consistently show 'things I didn't do' dominate end-of-life regret by 3-to-1 over 'things I did.'
- If the decision is reversible within 2 years, the regret-min frame is overkill. Use it for irreversible-within-5-years decisions.
- Your 80-year-old self has more time wisdom but less risk appetite. Weight both voices in the answer.
- This framework is BAD for decisions where you have moral obligations (to kids, to co-founders, etc.). Obligation trumps regret math.
- Don't run this when you're in an acute emotional state — breakup, firing, relationship crisis. Wait 2 weeks, then run it.
Customization tips
- Write out your answers to the three 80-year-old questions by hand before looking at the AI output. Your handwriting reveals what matters; typing flattens it.
- Run this framework 3 times: alone, with your partner, and with one person who knows you deeply. The three answers reveal the decision.
- If the framework output conflicts with your gut, notice that. Gut is valid data. Reconcile rather than overriding either.
- Set a re-decision date. If 'take this opportunity later' is the answer, calendar the decision point 12-18 months out. Otherwise 'later' becomes 'never.'
- Don't use this framework for decisions under $20k of lifetime consequence. Overkill. Use it for 5+ year irreversibles only.
Variants
Career-Pivot Mode
Specifically calibrated for leaving-a-job vs. staying decisions. Handles golden handcuffs and identity-bound-to-job dynamics.
Relationship-Decision Mode
For proposal, ending, or relocating-for decisions. Weights relational regret differently than career.
Later-Life Mode (50+)
For decisions after 50 where time horizons change the math. Action regrets compress, legacy considerations appear.
Frequently asked questions
How do I use the Regret Minimization Framework prompt?
Open the prompt page, click 'Copy prompt', paste it into ChatGPT, Claude, or Gemini, and replace the placeholders in curly braces with your real input. The prompt is also launchable directly in each model with one click.
Which AI model works best with Regret Minimization Framework?
Claude Sonnet 4.5 or Opus 4. Personal decision-weighting requires nuanced reasoning about life stages and trade-offs. Mid-tier and above.
Can I customize the Regret Minimization Framework prompt for my use case?
Yes — every Promptolis Original is designed to be customized. Key levers: Run this framework in writing. The act of writing the 80-year-old answer is 80% of the value; reading someone else's won't move you.; Inaction regrets compound. Studies consistently show 'things I didn't do' dominate end-of-life regret by 3-to-1 over 'things I did.'
Explore more Originals
Hand-crafted 2026-grade prompts that actually change how you work.
← All Promptolis Originals