As time goes by, AI technology has developed to offer us more and more comforts and conveniences in our daily lives. Continuing in this direction, Facebook in collaboration with Carnegie Mellon University (CMU), has developed an AI that can beat professional poker players.
The new AI system, called Pluribus, which was announced via a press release, managed to beat 12 of the best poker players in the popular six-player no-limit Texas Hold'em.
Pluribus played against 12 players in two games. In the first game, the AI played against 5 humans, and in the second, it was five AIs and one human. The AI bot was able to win an average of up to $1,000 per hour – an impressive amount for an AI.
The competition was attended by thousands of players over a two-day period.
The list of poker players includes Jimmy Chou, Seth Davies, Michael Gagliano, Anthony Gregg, Dong Kim, Jason Les, Linus Loeliger, Daniel McAulay, Nick Pietrangelo, Sean Ruane, Trevor Savage and Jake Toole.
It is believed that Pluribus was able to win the gameby applying Nash equilibrium theory which involves calculating equal probabilities of winning a game, as is done in the game of rock-paper-scissors.
Pluribus comes on the heels of a series of other such experiments, including Libratus AI, which was able to beat a human player when playing against an opponent. The new discovery suggests that AI can achieve superhuman abilities in new challenges.
How was Pluribus trained?
To be able to beat the top poker players, Pluribus was initially trained by playing with different versions of itself, so that it could learn to play from its trials and failures.
After a while, he was able to predict some of his opponent's moves, but not the entire game, by implementing a real-time search
It took just eight days to train the AI, using a 64-core server, less than 512 GB of RAM, and no GPUs. Furthermore, this AI is more affordable than similar ones.
Furthermore, Facebook and CMU hope to use the strategy employed by Pluribus' technology in areas such as cybersecurity , threat prevention , online auctions or navigation.
It may not seem like a great feat, but poker is one of the most difficult card games, as it involves bluffing and anticipating your opponent's moves, which even humans sometimes fail at. This ability of AI is something very important.
