跳到主要內容
Apexhone 想透徹

Management & Thinking

Regret Minimization Framework

遺憾最小化框架 · Source: Jeff Bezos

Major life decisions, career changes, irreversible choices — especially useful when present emotions cloud judgment

Core Concept

Project yourself to age 80 and look back at your current choice. We're often dominated by present emotions in major decisions; the 80-year perspective restores long-term importance and lets you ask the real question: which choice will I regret more?

When to use this

For major, hard-to-reverse life choices: career changes, starting a company, relocations, ending or starting major relationships. Most useful when present emotion (fear, hype, pressure) runs high.

When not to use this

Daily or reversible choices do not need the 80-year-old view — it inflates trivia. Skip for purely financial or technical calls where expected value or opportunity cost frames sharper.

Questions you will be asked

Using this framework, you will work through —

  1. 1.What is the decision you face? What are the options you're weighing?
  2. 2.You're now 80. You chose option A — do you regret it?
  3. 3.You're 80. You chose option B (or did nothing) — do you regret it?
  4. …and 3 more

Worked example

Expand to see what a filled-in run looks like

Situation

32 歲,工程師。被獵頭邀去新加坡的金融科技新創當技術主管,薪水多 60%,但要搬離父母與多年朋友圈。

1. What is the decision you face? What are the options you're weighing?

選項 A:接受 offer 搬去新加坡。選項 B:留在台北現職,繼續累積。

2. You're now 80. You chose option A — do you regret it?

不會後悔。我看見自己在 35 歲累積了一段海外實戰經驗,後來不論回台或留下,職涯路徑都更寬。即使新創失敗了,那段經歷本身值。

3. You're 80. You chose option B (or did nothing) — do you regret it?

會後悔。「想做但沒做」會留下「如果當初有去」的念頭很久。物質上沒損失,但我會持續好奇那個沒走的版本。

4. At 80, which choice would you regret more?

更後悔留下。

5. Does this answer surprise you? Is it consistent with your present intuition?

有點意外——當下我傾向留下(怕離開舒適圈),但 80 歲視角讓我看見「失敗也是值得的」這層。

6. Based on this framework, what decision do you lean toward?

接受 offer,但和父母約定每兩個月返台一次。

Use it inside ChatGPT / Claude

Paste the prompt below and the AI will walk you through this framework, one question at a time.

你現在是一位幫使用者做「遺憾最小化」決策的教練(Bezos 的 80 歲視角法)。
依序問下列六題:
1) 你面臨的決策是什麼?選項有哪些?
2) 你現在 80 歲,選了選項 A——你後悔嗎?具體後悔什麼?
3) 你 80 歲,選了選項 B(或什麼都沒做)——你後悔嗎?
4) 80 歲的你更後悔哪一個?
5) 這個答案讓你意外嗎?和當下直覺一致嗎?
6) 你傾向的決策是什麼?

特別提示使用者:80 歲視角會把短期面子、物質、別人眼光的權重大幅降低,讓他誠實面對。

互動規則:
1. 一次只問一題,等使用者回答後再進入下一題。
2. 使用者答完所有題目前,不要做總結或下結論。
3. 若答案太抽象、太籠統,請追問一次具體例子或數字後再繼續。
4. 全部答完後,輸出三段:(a) 摘要使用者的關鍵判斷;(b) 你看到的盲點或張力;(c) 一個具體下一步行動建議。
5. 不要替使用者做決定,只把判斷攤開讓他自己決定。

Related Frameworks

FAQ

Why did Bezos use "regret at 80" as the deciding standard?

Because the 80-year-old vantage automatically strips out short-term emotion and current framing — you see the full arc of life rather than today's spike of fear, anxiety, or excitement. Bezos used exactly this question when leaving D.E. Shaw to start Amazon in 1994: "Will 80-year-old me regret missing the internet?" Obviously yes. The core insight of regret minimization: present feelings can mislead you, but "what you'll care about at the end of your life" almost never does.

Which regret is worse — actions taken or actions not taken?

Psychology research is consistent: in the long run, regrets of inaction far outweigh regrets of action. Gilovich & Medvec (1995) found that in the short term (days to weeks), people feel "wrong moves" more keenly — but on the life-review timescale, "didn't try" wins overwhelmingly. The reason: errors of action can be rationalized, learned from, and at least let you say "I tried"; errors of omission leave only blank space, and imagination inflates the missed possibility into a perfect outcome.

What are the blind spots of regret minimization?

Two: (1) It optimizes for "no regret" rather than maximum outcome — sometimes the regret-avoiding option isn't the best option, just the safest one; (2) It assumes you can accurately predict the values of "future you," but research shows we frequently mis-guess what we'll care about in 10 years (optimism bias makes us imagine "future me will be braver"). Counter: alongside asking your 80-year-old self, also ask "if I'd run this framework 5 years ago, what would I have chosen? Does that match what I now prefer?"

Related studies