AI Wealth Truth (02): Why Can Randomness Create Extreme Inequality?

I. Assume a perfectly fair world. Everyone has the same ability. Everyone starts at the same point. No discrimination. No bias. No unfairness. In this world, does wealth inequality disappear?

II. No. The gap grows larger and larger. This is not conspiracy. It is math.

III. In 2018, Italian physicist Alessandro Pluchino and his team ran a computer simulation. They simulated the 40-year careers of 1,000 people. Each person's "talent" was randomly assigned from a normal distribution. Most people had average talent, a few had very high or very low talent.

IV. Then they made each person experience an "event" every six months. Events were random. 50% were "lucky" and 50% were "unlucky". If you hit a lucky event, your wealth doubled. If you hit an unlucky event, your wealth was cut in half. Note: this had nothing to do with talent. It was pure luck.

V. After 40 years (80 events), what happened? The richest 20% owned more than 80% of the wealth. Almost a perfect match to the Pareto pattern in the real world. And the richest people were often not the most talented. They were simply the luckiest.

VI. How is that possible when the rules are fair? The problem is multiplication.

VII. If returns were additive, luck would average out. You gain 100 today and lose 100 tomorrow. It cancels. In the long run, wealth should converge toward the mean.

VIII. But if returns are multiplicative, luck is amplified. You double today and halve tomorrow. It looks like it cancels. It does not.

IX. Suppose you have 100. Year one: double to 200. Year two: halve to 100. Net change: 0.

X. Now reverse the order. Suppose you have 100. Year one: halve to 50. Year two: double to 100. Net change: still 0.

XI. Looks fair. Now consider streaks. Two good years versus two bad years.

XII. Player A gets lucky twice: 100 -> 200 -> 400. Player B gets unlucky twice: 100 -> 50 -> 25. Gap: 400 vs 25. A 16x difference. The only difference is the sequence of luck.

XIII. What if you run this for 40 years? A 40-year lucky streak is unlikely, but someone will get close. A 40-year unlucky streak is also unlikely, but someone will get close. Extreme cases shape the distribution.

XIV. That is the nature of a multiplicative random process. In a multiplicative world, luck is not noise that gets averaged out. Luck accumulates and amplifies. Good luck stacked on good luck becomes extreme wealth. Bad luck stacked on bad luck becomes poverty.

XV. Real-world wealth growth is exactly multiplicative. If you have 1 million and earn 10%, you gain 100,000. If you have 10 million and earn 10%, you gain 1,000,000. Same return rate. Ten times the absolute gain. The higher the base, the larger the multiplicative effect.

XVI. More brutal: loss is multiplicative too. If you have 1 million and lose 50%, you are left with 500,000. You need a 100% gain to get back to 1 million. If you have 10 million and lose 50%, you are left with 5 million. You also need a 100% gain to get back. In a multiplicative world, digging the hole is easier than filling it.

XVII. Pluchino's simulation revealed something worse: The most talented people are often not the richest. Talent affects how efficiently you exploit good luck. But good luck itself is random. Average talent plus explosive luck beats high talent plus ordinary luck.

XVIII. This attacks every success narrative we have. We assume winners must have something special. "They are smarter." "They work harder." "They know a secret we do not."

XIX. No. They might simply be luckier. And luck is not replicable.

XX. This is why "success formulas" rarely work. You copy everything a winner did. You read their book, listen to their talks, replicate their habits. Can you replicate their luck? No. In a multiplicative world, luck is the decisive variable.

XXI. AI makes this multiplicative effect more extreme. In the old world, even with good luck, productivity had a ceiling. You have 24 hours a day and 365 days a year. In the AI era, productivity can be copied. You write an AI program and it can serve a million people at once. Your single "lucky hit" gets amplified a million times.

XXII. This means that in the AI era, the role of luck is further magnified. Before, the lucky might end up 10x richer than the unlucky. Now the lucky might end up 10,000x richer. The multiplier is bigger, so the gap becomes more extreme.

XXIII. This is not to say effort is useless. Effort moves you into a position where you can exploit luck. If you do not enter the game, you cannot catch a lucky break. But once you enter, outcomes depend heavily on luck. Effort is the entry ticket. Luck is the allocation ticket.

XXIV. Pluchino's paper title is telling: "Talent and Luck: The role of randomness in success and failure." Their conclusion: we should rethink how resources are allocated. If success is largely luck, progressive taxation and redistribution gain a new moral foundation.

XXV. But before policy changes, you need this truth: If you did not succeed, it might not be because you are not smart enough. It might not be because you did not work hard enough. It might be because the multiplicative random process did not land on your side. That is not your fault. That is math. AI-era math makes it harsher.