Meta
First, think hard about whether a decision is a one-way door (irreversible, or reversible only with high costs) or a two-way door (reversible)
Decisions that are one-way doors include
- Working on something with clauses attached (NDA, non-compete, lockin period etc)
- Long-term contracts with clients and vendors
- Who to date seriously or marry (in the former, damage from bad choices is mostly emotional. In the latter it's emotional and financial)
- Where to live
- Decisions about health and anything that involves risk of ruin
Decisions that are two-way doors include
- What to work on without clauses attached (no crazy NDA, no lockin period etc)
- Casual dating
- Where to travel to, what to wear, what to spend your time in the short-run etc
Spend a lot of time thinking about the one-way doors. But if it's a two way door, just roll with it without thinking too much. Remember that Action Produces Information
Making irreversible decisions
When making irreversible decisions, ask yourself the following:
Why I should say no
- Does this have a risk of ruin?
- Does this limit my long-term upside from life in a significant way?
- Specifically, does it have harmful NDAs or non-competes?
- Does this have very high opportunity cost?
- Does this take too much of my time for an elongated period, and limit the experiments I can do?
- Does this drain too much of my emotional, creative, or intellectual energy?
Why I should say yes
- Does this have a high potential upside?
- Does this help me in the following areas:
- It makes me wealthier, accounting for opportunity cost
- It gives me more optionality and more control of my time
- It opens new doors for me (human network)
- It uplifts me emotionally (in a sustained way — not a spike and crash)
- Does this help me get better at skills I want to develop further?
Making reversible decisions
Don't overthink it. Remember that action produces information. Just try things out, reflect on learnings, and either double down or move on.
It's difficult to get into this habit. But do it. Think of it as stochastic gradient descent. If you keep trying, reflecting on what you learnt, and changing to iterate quickly — you'll end up at the optimal way of doing things sooner rather than later

Summary of Seeking Wisdom: From Darwin to Munger (how I arrived at the heuristics above)
<aside>
💡 The full summary and highlights are at Seeking Wisdom: From Darwin to Munger
</aside>
We should be aware of our biases, and should also learn why people make bad decisions. Then, we should codify these learnings into checklists so that we don't ourselves make bad decisions. To do this, we should use mental models and look at evidence dispassionately — while being aware of our cognitive biases.
Remember that you only need to get rich once. Figure out what your goal is and how to get there. And don't take stupid risks that could lead to ruin.
First, figure out what has worked well in the past. Ask:
- What was my original reason for doing something? What did I known and what were my assumptions? What were my alternatives at the time?