Regret Minimization Framework

Where this came from

In 1994, Jeff Bezos was working as a senior vice-president at D.E. Shaw, one of the most respected quantitative hedge funds in the world. He came across a statistic that internet usage was growing at 2,300 percent annually and became convinced that online commerce would become significant. The problem was that he had a well-paying, prestigious job and no certainty that an online bookstore would succeed.

Bezos described his reasoning process in a 2010 commencement address at Princeton. He said he projected himself forward to age 80 and asked which choice he would regret more: having tried and failed, or never having tried at all. The answer was clear. He would not regret failure. He would regret never having made the attempt. He resigned from D.E. Shaw, drove cross-country with his wife MacKenzie, and wrote the Amazon business plan during the drive. The company launched in 1995 from a garage in Bellevue, Washington.

Bezos has cited this framework repeatedly since then, describing it as his standard tool for evaluating major personal and professional decisions. The core insight is simple: regret over inaction compounds more than regret over action, because inaction leaves you permanently wondering what might have been. You can apply the regret minimisation framework to your decision with AI guidance using the same three-step sequence Bezos described.

How it works

The framework has three steps. First, identify the decision you are facing and the options available. Second, imagine yourself at age 80, in a reflective state, looking back at your life. Third, ask: if I choose Option A, will I regret not having chosen Option B? If I choose Option B, will I regret not having chosen Option A?

The 80-year-old vantage point matters because it strips away the distractions that dominate present-moment decision-making: what colleagues will think, how uncomfortable the transition period will feel, whether the timing is right. From 80, those concerns look small. What remains is whether the life you built reflected what you actually valued.

The framework is particularly powerful for decisions involving action versus inaction. Research by psychologists Thomas Gilovich and Victoria Medvec (1994) found that people more often regret things they did not do than things they did do, particularly over long time horizons. In the short term, regrets of action dominate. Over a lifetime, regrets of inaction take over. The framework essentially asks you to weight the long-term version of regret more heavily than the short-term version.

One practical constraint: the framework requires that you can form a meaningful picture of what you would regret. For decisions where you have no basis to project into the future, for instance, choosing between two technical career paths in a field that may not exist in its current form in 10 years, the 80-year-old self has nothing reliable to consult. In those cases, supplementary tools are needed.

When to use it and when not to

Use the Regret Minimization Framework for decisions that are hard to reverse, involve a significant departure from your current path, and have consequences that compound over years. Career pivots, starting a company, moving to another country, committing to or ending a long-term relationship. These are choices where the long-term horizon is the most relevant one.

Do not use it for decisions that are easily reversible, low-stakes, or primarily short-horizon. Choosing which marketing agency to hire, whether to accept a speaking engagement, or how to price a product do not benefit from imagining your 80-year-old self. The framework adds overhead without adding insight.

One important limitation: humans are poor predictors of their future emotional states. Psychologist Daniel Gilbert's research on affective forecasting shows that we systematically overestimate how bad we will feel about failures and underestimate how quickly we adapt. This means the regrets you project at 35 may not match the regrets you actually experience at 80. Use the framework as a directional signal, not a precise calculation. It tells you which way to lean, not exactly how much you will feel it.