Computer programs have shown superiority over humans in two-player games such as chess, Go, and heads-up, no-limit Texas hold'em poker. However, poker games usually include six players--a much trickier challenge for artificial intelligence than the two-player variant. Brown and Sandholm developed a program, dubbed Pluribus, that learned how to play six-player no-limit Texas hold'em by playing against five copies of itself (see the Perspective by Blair and Saffidine). When pitted against five elite professional poker players, or with five copies of Pluribus playing against one professional, the computer performed significantly better than humans over the course of 10,000 hands of poker. Science, this issue p. 885; see also p. 864
A new paper published by Science today details how Libratus, an AI developed by researchers at Carnegie Mellon's computer science department, managed to take on and defeat top industry professionals at one of the most challenging forms of poker: No-limit Texas Hold'em. Yes, the same variant of poker that swept the nation during the heady days of the early to mid-aughts.
This talk gives a high-level explanation of Libratus, the first AI to defeat top humans in no-limit poker. A paper on the AI was published in Science in 2017. No-limit Texas hold'em is the most popular form of poker. Despite AI successes in perfect-information games, the private information and massive game tree have made no-limit poker difficult to tackle. We present Libratus, an AI that, in a 120,000-hand competition, defeated four top human specialist professionals in heads-up no-limit Texas hold'em, the leading benchmark and long-standing challenge problem in imperfect-information game solving.
In a landmark achievement for artificial intelligence, a poker bot developed by researchers in Canada and the Czech Republic has defeated several professional players in one-on-one games of no-limit Texas hold'em poker. Perhaps most interestingly, the academics behind the work say their program overcame its human opponents by using an approximation approach that they compare to "gut feeling." "If correct, this is indeed a significant advance in game-playing AI," says Michael Wellman, a professor at the University of Michigan who specializes in game theory and AI. "First, it achieves a major milestone (beating poker professionals) in a game of prominent interest. Second, it brings together several novel ideas, which together support an exciting approach for imperfect-information games."
Superhuman performance by artificial intelligence (AI) has been demonstrated in two-player, deterministic, zero-sum, perfect-information games (1) such as chess, checkers (2), Hex, and Go (3). Research using AI has broadened to include games with challenging attributes such as randomness, multiple players, or imperfect information. Randomness is a feature of dice games, and card games include the additional complexity that each player sees some cards that are hidden from others. These aspects more closely resemble real-world situations, and this research may thus lead to algorithms with wider applicability. On page 885 of this issue, Brown and Sandholm (4) show that a new computer player called Pluribus exceeds human performance for six-player Texas hold'em poker.