Friday September 06,2013 : LATEST ADVANCE IN ARTIFICIAL INTELLIGENCE MAKES THIS POKER BOT UNBEATABLE
Norwegian neural network technology behind latest generation robot
The New York Times reports that the latest developments in poker Artificial Intelligence have achieved unprecedented levels of robot experience-learning to produce an almost unbeatable poker bot for research and possibly gambling purposes.
Poker-playing robots have been hitting the headlines for years as scientists and academics have used the game to test their AI work against expert human players, with recent contests showing a growing level of smarts in the machines, enabling them to learn from their mistakes and make the correct moves next time around.
The latest developments have an interesting background, involving Norwegian scientist Fredrik Dahl, who at one time worked for the Norwegian Defense Research Establishment in Kjeller, Norway developing artificial intelligence for secret government projects on combat simulations.
"The job involved using neural networks," the New York Times reports. "Functioning much like an extremely focused, one-dimensional version of the human brain, these complex computer algorithms develop strategies that emerge through so many repetitive mathematical calculations that few humans could reproduce, much less endure them. Dahl’s work on two-sided, zero-sum games, where there is no mutual interest, proved to be useful in developing strategies to win not only wars but also poker games."
The article reveals that sophisticated poker bots use knowledge gained from billions of staged rounds of poker fed through neural networks, and the result is an unpredictable poker player that can win almost every time.
Three different banks of knowledge are used by the bot to prevent defeat.
"The theory behind it is almost paranoid," Dahl explains told the Times. "It's estimated that only 100 players around the world even have a chance of taking the game down."
Gamblers might win a given hand out of sheer luck, but over an extended period, as the impact of luck evens out, they must overcome carefully trained neural nets that self-learned to play aggressively and unpredictably with the expertise of a skilled professional.
Poker is primarily a game of skill and intuition, of bluffs and traps in which the player tries to play his opponent as well as his cards, the article reports.
"This machine does that, responding to opponents’ moves and pursuing optimal strategies. But to compete at the highest levels and beat the best human players, the approach must be impeccable."
The piece goes on to recount the experiences of G2 Game Design chief Gregg Giuffria, whose company developed Texas Hold ‘Em Heads Up Poker.
Whilst testing a prototype of the program he thought he detected a flaw. When he played passively until a hand’s very last card was dealt and then suddenly made a bet, the program folded rather than match his bet and risk losing more money.
“I called in all my employees and told them that there’s a problem,” he says. The software seemed to play in an easily exploitable pattern. “Then I played 200 more hands, and [the robot] never did anything like that again. That was the point when we nicknamed him Little Bastard.”