One way door decisions

There are moments in life when you hit slow-burn-max mode, when you know a big decision is coming, and you can feel the weight of it. You stop everything else and think deeply about the problem you’re about to face.

Jeff Bezos calls these “one-way doors.” Most decisions are two-way doors, you can go through, try it out, and walk back if it doesn’t work. But some aren’t like that. They’re irreversible, or feel that way from where you’re standing. These decisions look different for everyone, but for me, they’ve shown up at inflection points: moving countries, choosing a first job, getting married.

You could argue none of these are truly irreversible. But zoom out to the perspective of a full lifetime, say, 100 years, and the cost of reversing such choices starts to feel steep. By 30, you’re probably halfway through your productive output. So these one-way doors feel even harder. There’s no silver bullet. The best you get is a series of tradeoffs—some heavy on one side, some on the other.

I still struggle with these decisions. I’m not claiming mastery. But I do have a general approach that helps me move forward when I’m staring down one of these one-way doors.

I start with “explore” mode. I gather inputs, talk to people, look for frameworks, test small hypotheses. Only once I’m confident that I’ve mapped enough of the space do I switch to “exploit” mode—where I narrow down and commit.

This mindset mirrors the secretary problem:

You’re interviewing candidates one by one, in random order, for a single role. After each interview, you must decide on the spot—hire or move on. You can’t go back. The optimal strategy is to reject the first 37% of candidates to gather a baseline, and then hire the next one who’s better than everyone you’ve seen so far. This gives you the best chance of hiring the top candidate.

I use the same logic when seeking opinions or inputs. If I plan to speak to 20 people before making a call, I treat the first 7 as pure exploration. I gather everything, say yes to all the ideas, take detailed notes—but I don’t commit. Only after that first phase do I shift into selection mode, looking for the best fit that exceeds the baseline I’ve now formed.

You never really know what you don’t know. And you can’t learn everything, time’s limited. So this 37% rule offers a structure. If I’ve got 6 months before a major decision, I spend the first 1.5 months on deep exploration. I talk to people with different lived experiences and world views. I soak in their pros and cons, their assumptions, their logic.

And when I hear a new viewpoint, I try to hold it as “true,” just for a while. I let it challenge whatever internal direction I’ve started drifting toward. This takes effort. The temptation is to build a fortress around early ideas—to protect them from challenge. But the harder (and better) path is to seek disconfirmatory evidence. Poke holes in your own arguments before someone else does.

As you gather more input, you’ll start to feel a tension between conflicting views. That’s a good thing. You can stretch your thinking, add nuance, let your stance evolve.

Eventually, after enough of these conversations—human or otherwise (ChatGPT, Perplexity, Deepseek, etc.), you arrive at something stable. You marinate on it. For me, that settling phase takes 4–7 days. The emotional churn quiets down, and something clear starts to take shape.