An interesting tournament is currently on-going at Pittsburgh's Rivers Casino, where four of the world's best casino players will compete against a computer program developed by Carnegie Mellon University (CMU). This is a rematch to determine if Carnegie Mellon's AI can really beat humans on one of the world's toughest games, the poker.
The tournament is being heralded as 'Brains Versus Artificial Intelligence', and so far, CMU's AI program called Libratus, is the chip leader. However, the players and AI's creators reminded the public that it's still too early to make a verdict about who's going to win the tournament, considering that the game has just begun and it is a 20-day competition.
In 2015, the result of the competition was reversed, where three of four humans came on the top and Carnegie Mellon's AI landed on fourth place for collecting fewer chips. But the results is not just for the sake of a successful science experiment as this type of AI can be used in complex scenarios, not just in a game of poker where decisions need to be executed in situations where information is inadequate or misleading.
For this event, CMU School of Computer Science partnered with Rivers Casino, Pittsburgh Supercomputing Center (PSC), and Sandholm's Electronic Marketplaces Laboratory. The poker tournament is sponsored by GreatPoint Ventures, Avenue4Analytics, Intel and Optimized Markets Inc., TNG Technology Consulting GmbH, and the journal Artificial Intelligence.
Started last Jan. 11, professional poker players will play 120,000 hands of Heads-Up No-Limit Texas Hold'em against a CMU computer program for a $200,000 prize purse. The human players, Jason Les, Dong Kim, Daniel McAulay, and Jimmy Chou, are looking forward to defeating CMU's improved Claudico.
For this rematch, the human players agreed to increase the number of hands played to decrease the possibility that neither the humans or the computer will win through pure luck. That said, the humans will need to play more days and play two hands simultaneously.