He used AI for trading and earned $9.8 billion in a year, averaging $27 million per day. He ranked 20th on the UK Rich List last year and has risen to 8th this year This person's name is Alex Gerko, 46 years old, with a PhD in mathematics. There is not a single trader in the company AI only does one thing with him, guess if the price will go up or down in the next few seconds to ten minutes It sounds simple, but this matter is extremely difficult because before the price changes, there will be a bunch of clues left in the market, such as the speed of orders coming in, the proportion of buying and selling orders, how much goods are listed on a certain price, and who is canceling orders. The human brain has no time to see, and AI can simultaneously monitor tens of thousands of product types Guess exactly how he turned it into money. He put a buying price and a selling price on each variety at the same time. The model said it would rise next, so he put the buying price higher and the selling price lower. The model said it would fall, so he did the opposite After the transaction, what he earns is the gap between the buying price and the selling price, but he also has a pile of extra goods in his hands. It's okay if they rise, but if they fall, he has to find a way to deal with them immediately. Either hedge elsewhere, or turn the quote in the opposite direction and ask the next person to quickly take it away The true fate of this profession is another matter. The price you quoted is often taken away by others because you have just received news and your reaction is slow. The person who eats you knows earlier than you, and after finishing the price, they immediately move in the direction of your loss So who can survive depends on whether the model is slightly more accurate than the opponent on average. It doesn't matter if it's not very accurate. He makes tens of millions of transactions a day, and multiplying that by tens of millions of transactions is billions of dollars His strength lies in the fact that one model can cover 53000 varieties, including stocks, foreign exchange, bonds, gold, crude oil, and cryptocurrencies, sharing a set of predictive capabilities. With one advancement in the model, he can make more money in all markets at the same time. This is why the company's 300 employees can earn so much money In order to keep the model updated and data readily available, he decided to build his own computer room. Iceland's one uses geothermal power, has over 25000 graphics cards, 650PB of data, and Finland will spend $1.1 billion to build five more The more chaotic the market, the more he earns. When prices jump up and down, everyone is more eager to find someone to immediately make a deal. His price difference is widened, and the trading volume doubles In the end, all the money belongs to him, without external investors or helping others manage it. He will earn $1.7 billion in 2024, with 31 partners sharing it, and he alone will receive $920 million There are four things that ordinary people can learn from 1/Let the machine do repetitive tasks, leaving only the judgment step for oneself 2/After collecting the toll, don't bet on the direction. Look for things that others are in a hurry to do and can only ask you to do. Earn a little at a time and earn him ten thousand times 3/The small advantage of being able to verify right and wrong on the same day, don't wait for three years to know if it's feasible 4/Minimizing costs, building your own house is more cost-effective than renting, the price war ends in the end, and those who survive are the ones with the lowest costs