AI Wealth Truth (05): Why the Role of Luck Is Systematically Underestimated by 90%

I. In 2012, two Italian physicists ran a simulation experiment. They created a virtual world with 1,000 people, each with a "talent score". Talent was randomly assigned from a normal distribution. Most people had average talent, and a few had extremely high or extremely low talent. Then they let these virtual people live through 40-year careers.

II. Every six months, each person encountered a random event. If it was a "lucky event" (50% probability), your wealth doubled. If it was an "unlucky event" (50% probability), your wealth was cut in half. The events were unrelated to talent. Pure randomness. Talent mattered only in this way: when a lucky event happened, high-talent people could exploit it better.

III. After 40 years, what happened? The richest people were not the most talented. Their talent was only slightly above average. They became the richest because they had the best luck, hitting more lucky events.

IV. Even more striking: the correlation between talent and final wealth was very weak. Among the most talented group, wealth outcomes were widely dispersed. Some became very rich, but many stayed in the middle, and some ended up poor. Why? Because the variance of luck was so large that it overwhelmed the effect of talent.

V. This simulation is known as the "Talent vs Luck Model". It reveals a brutal fact: In a multiplicative growth system, the role of luck is systematically underestimated. And the role of ability is systematically overestimated.

VI. Why do we underestimate luck? The first reason is attribution bias.

VII. Winners have a strong motive to attribute success to ability. If you admit "I was mostly lucky", your self-esteem takes a hit. You also cannot publicly claim you "deserve" the success. So winners instinctively build a narrative: I worked hard, I had vision, I persisted, I am different. Winners believe this narrative themselves. This is not deception. It is self-protection.

VIII. Psychologists have surveyed lottery winners. Many winners later recall: "I had a feeling I would win." "I chose numbers strategically. Not randomly." They assign a fake sense of "agency" to pure luck. The brain cannot accept the fact: "I had no control at all."

IX. The second reason is survivorship bias.

X. We can only observe the winners. Failures do not write autobiographies, appear on talk shows, or become business school cases. When we study "the secret of success", we study only the successful. In statistics, this is selection bias.

XI. Suppose 1,000 people all start companies using the same method. 990 fail. 10 succeed. We interview the 10 winners and ask: "What is your secret?" They say: "I believed in my product", "I persisted", "I dared to take risks". But the 990 failures also believed, persisted, and took risks. The difference is not the method. It is luck.

XII. Warren Buffett is one of the most successful investors in history. His method is studied and copied by countless people. But few mention this: he was born in America in 1930 and rode a century of rapid U.S. growth. If he had been born in Argentina in 1930, Argentina's stock market fell more than 90% during his investing lifetime. The same "value investing" in Argentina would have bankrupted him. How much of Buffett's success is method, and how much is the luck of birthplace?

XIII. Bill Gates founded Microsoft and became the world's richest man. His story is told endlessly: dropping out, taking risks, technical genius. But few mention this: his mother was friends with an IBM board member. That relationship led IBM to sign the operating system licensing deal that changed history. Without that "luck", there might be no Microsoft today. Ability let Gates seize the opportunity, but the opportunity itself came from networks and luck.

XIV. Spotify founder Daniel Ek was born in Sweden. Sweden has free higher education, strong social safety nets, and high-speed internet infrastructure. He could focus on building a company without fearing that failure would leave him on the street. If he had been born in a country without safety nets, he might not dare to take the risk. Environmental luck is the hidden prerequisite of success.

XV. The third reason is the narrative fallacy.

XVI. The human brain is a story machine. We cannot accept "things just happen randomly." We must assign causality, meaning, and pattern. "He succeeded because he worked hard." That is a good story. "He succeeded mostly because of luck." That is not a good story. We choose good stories, not the truth.

XVII. Nassim Taleb used a metaphor in The Black Swan. Imagine a turkey that gets fed by the farmer every day. From experience it "learns": the farmer is a good person and will feed me every day. It becomes more and more confident in this "law." Until Thanksgiving, when it gets slaughtered. The turkey's "success experience" cannot predict the decisive black swan event.

XVIII. Many winners are like that turkey. Their methods worked in the past, mainly because the environment happened to favor them. They conclude: my method is correct. Until the environment changes and their method suddenly fails. Success experience may be only a fossil of past luck.

XIX. The AI era makes this problem sharper. AI startup success cases are widely promoted. "They used AI to do X, and they succeeded." But we do not see the countless cases that also used AI to do X and failed. Survivorship bias is extremely severe in AI.

XX. And AI's own success contains a large luck component. Why did OpenAI succeed? Partly because they happened to get enough compute at the right time. Happened to recruit key researchers. Happened to choose the Transformer architecture. If any one of these "happened to" did not occur, there might be no GPT today.

XXI. AI amplifies luck. In the traditional world, luck fluctuations are smoothed by time and effort to some extent. In the AI world, a lucky event can be copied and amplified without limit. The first person to build a breakout AI product might take the whole market. Second place might get nothing. The decisive power of luck reaches a historical peak in the AI era.

XXII. What does this mean for you? First, stop blindly imitating winners. The "success secrets" they teach you may just be their misreading of their own luck. Copying their methods will not give you the same luck.

XXIII. Second, stop blaming yourself. If you worked hard but did not succeed, it may not be because you are not smart enough or not hardworking enough. It may simply be that random events did not land on your side. That is not your fault.

XXIV. Third, increase your number of "entries". If success is largely luck, increase your chances of meeting luck. Try more. Place more bets. Do not put all your chips on one road. You cannot control the dice. But you can control how many times you roll.

XXV. The role of luck is systematically underestimated. Not by 10%. Not by 30%. Maybe by 60%, even 90%. Winners will not tell you this truth, because they do not know it themselves. They are blinded by attribution bias. We are blinded by survivorship bias. In the AI era, the weight of luck will only grow. This is not pessimism. It is probability theory.