I was one of those who said it could never be done: that a computer wouldn’t ever manage to beat the best at the game of poker. I was romantic and wide-eyed at 18, when my heroes were the cowboys from Texas who ruled the felt. They were uneducated and coarse, yet chock full of the human qualities needed to excel at poker.
With nicknames like Amarillo Slim and Texas Dolly, these larger-than-life characters had fearlessness, aptitude, and a deep understanding of what makes people tick. The higher the stakes, the better they played. One moment an opponent would be confidently betting their hand, then squirming with fear and unable to call.
The game they played was No Limit Texas Hold’em, often called the Cadillac of all poker games, because every hand comes down to a brutally simple question of “is he bluffing or is he not?” How could a computer evaluate that?
The cowboys gave way to the poker savants. In the 1990s my hero was Phil Hellmuth, the “Poker Brat”, a Wisconsin-born, pimply-faced kid who used what he called his “White Magic” to become the most decorated poker player ever. Hellmuth could read your soul, it was said, as soon as he heard you speak. A wrong-way glance, lying eyes, or a trembling hand were to Hellmuth much the same as peeking at your cards.
But the internet changed poker for good. Online players saw more hands of poker in one year than most cowboys had played in a lifetime. Millions more hands, in fact. Reams of data. And what they found, using computer software, is that the binary patterns of a person’s bets and calls, of every action they take on the table, gave as much insight into whether someone was bluffing or true than even a hotline to the soul. We would all like to believe that humans are random, capable of doing anything at any time. The data shows different. When they first started coming into the live arena, the old pros laughed at these basement kids. They wore sunglasses to cover their eyes, hooded sweatshirts to shield their face, and scarves to hide a pulsating neck. Internet-trained players were mummified robots who didn’t say a word, but what they did do is begin to win.
And what separated the best from the rest was how they used computers to improve their play, to analyse and predict from mountains of data.
In fact if you met the group of people who represent the top poker players in the world today, you might say that this group of people are closer to computers than any group of people you have ever met. Devoid of emotion, ruled by logic, they have the ability to quickly process masses of data, make decisions and execute them. They are the best guessers in the history of the world, and there is no magic about it.
Last month it was announced that a computer program had beaten the world’s best at poker. We long assumed that, to achieve this, computers needed to become more like humans. But in fact this story is about humans becoming more like computers. This is about us realising that little computers is maybe all we are.
I will always treasure the artist, the swashbuckler, the bohemian genius in the writers’ room with long hair and a quick wit. I’ll always pull for the one who seems to be able to do it with no rules at all. But if poker is any guide, computers will help do all of those jobs better until they do them best. And they won’t do it by becoming more like us. It’ll happen when we realise that we’re more like them. That has been the story of poker.
Source: The Guardian
