AI Wealth Truth (25): Why Sunk Costs Drain Your Wealth
The sunk cost fallacy: giving up what you already invested triggers pain. The brain would rather keep losing than "admit defeat"
I. You spend 1,000 on a concert ticket. On the day of the show, you get sick and feel terrible. Do you still go? Many people force themselves to go. Their reason is: "I paid 1,000. If I do not go, I lose it."
II. But that reason is wrong. That 1,000 is already gone. Whether you go or not, it will not come back. The real choice is: A. Go while sick and suffer through a painful concert. B. Stay home, rest, and recover. B is clearly better. But the brain pushes you toward A.
III. This is the sunk cost fallacy. Costs already paid should not affect future decisions. But the brain cannot do that. The brain treats "giving up what you already put in" as a loss, and loss hurts.
IV. Evolutionarily, this made sense. In times of scarcity, giving up what you had in hand was a serious loss. But in modern decision-making, the instinct becomes a trap.
V. The sunk cost fallacy is especially deadly in investing.
VI. Example 1: clinging to losing stocks. You spend 50,000 on a stock. It is now worth 20,000. You refuse to sell, because "if I sell, I lose 30,000." You wait for it to "come back". But that 30,000 is already lost. Whether you sell or not, it will not return. The real question is: if you had 20,000 in cash today, would you buy this stock? If the answer is no, you should sell. Capital should go to the best use, not to "waiting to break even".
VII. Example 2: continuing to fund a failing project. You start a business and put in 500,000. The company is losing money. Rational analysis says you should cut losses. But you think: "I already put in 500,000. If I quit now, it is all wasted." So you invest more, and lose 1,000,000. Sunk costs make people sink deeper and deeper.
VIII. Example 3: sticking with the wrong career path. You studied a major for five years and worked in the industry for ten. You realize it does not fit you and want to switch. But you think: "I invested 15 years. Switching is such a waste." So you keep going. Ten years later, it is even harder to switch. You use future life to pay for past decisions.
IX. Example 4: staying in an unhappy relationship. "We have been together for five years. Breaking up now would be such a waste." But those five years are already gone. The real question is: how do you want to live the next five years? Sunk costs trap people inside unhappy relationships.
X. Why is the sunk cost fallacy so hard to overcome?
XI. Reason 1: loss aversion. Admitting sunk costs means admitting a loss. Loss is painful. The brain avoids pain by default. Continuing lets you delay the pain of confirming the loss. It is an emotional ostrich strategy.
XII. Reason 2: the need for consistency. Humans want to feel consistent. Admitting your past decision was wrong creates cognitive dissonance. Continuing preserves the illusion that "I was right back then". Staying consistent with your past feels more comfortable than making the right decision.
XIII. Reason 3: other people's judgment. Quitting a project you invested in may invite ridicule. "He finally admitted failure." Continuing helps you save face, at least for now. Social pressure amplifies the sunk cost fallacy.
XIV. In the AI era, the sunk cost fallacy is exploited systematically. Subscription services want you not to cancel. So they design "accumulated points" and "achieved tiers". Those are sunk costs. "I am already a gold member. Canceling would be a waste." The only purpose of those points and tiers is to trap you.
XV. Games use the same mechanism. You already spent 1,000 hours in the game. You already spent thousands buying gear. Quitting feels like a waste. So you keep spending time and money. Game designers understand sunk cost psychology precisely.
XVI. Learning software or tools creates similar lock-in. "I already learned this tool. Switching is too much work." Even if a new tool is clearly better, you resist. This is technical sunk cost lock-in.
XVII. How do you overcome the sunk cost fallacy?
XVIII. 1. Ask: "What if I started from zero?" Forget what you already invested. Imagine you start today with no history, no baggage. Would you make the same choice? If not, you should quit. This thought experiment helps you bypass sunk costs.
XIX. 2. Separate past from future. The past is past. Its only value is the lesson. Decisions should be based entirely on future costs and future benefits. Do not treat "past investment" as an input to "future choice". Time only moves forward.
XX. 3. Write down opportunity costs. What else could the time and money you keep investing be used for? Write the alternatives down. When you see "I could use this capital elsewhere", quitting becomes easier. Make opportunity cost visible.
XXI. 4. Accept loss as part of growth. Everyone makes bad decisions. Admit the mistake and cut losses. That is smarter than insisting until bankruptcy. Failure is not shameful. Refusing to face failure is.
XXII. 5. Set stop rules. Before you start any project, define exit conditions. "If it is not profitable in six months, shut it down." "If it drops 30% below cost, sell." Use preset rules instead of on-the-spot judgment. Rules can overpower emotion.
XXIII. Sunk costs are a ghost. They do not exist in the future. They exist only in the past. Yet they shape your future decisions. Letting a ghost decide your life is absurd.
XXIV. You already read most of this essay. If you think it is useless, will you still finish it? If yes, you might be trapped by sunk cost. The right move is: evaluate only the value of continuing, not the time you already spent reading. In the AI era, sunk-cost traps are designed more precisely. But the principle does not change: the past should not kidnap the future.
AI Wealth Truth (24): Why Your Intuition About Low-Probability Events Is Catastrophically Wrong
The probability weighting function: the brain turns 1% into 5%, and 99% into 80%
AI Wealth Truth (26): Why You Overpay for "Optionality"
Option value bias: we overvalue future flexibility and get harvested by options and installment plans
AI Practice Knowledge Base