Imagine going into a cavernous casino showroom, sitting down at a poker table, and playing against a computer.
Who do you think would win?
There have been past experiments with computers vs. chess players and even people using computers t win at blackjack. Results have been varied over the years but a recent experiment with a computer designed by Carnegie Mellon University was an eye-opener for the poker world.
The computer, known as Libratus, was designed by computer scientist Thomas Sandholm and PHD student Noam Brown. Four professional poker players accepted the challenge to try to beat Libratus.
During the session called 'Brains Vs. AI,' a total of 120,000 hands of no-limit poker were played under normal poker rules regarding calls, raising or folds. At the end of the session, the computer had piled up 1,776,250, soundly defeating its human opponents.
The poker pros were Jimmy Chou, Jason Less, Daniel McAulay and Doug Kim, each of whom started with 20,000 in chips. The well publicized game was held at Rivers Casino in Pittsburgh, PA.
While the players admitted they expected the computer would win, they thought the game would be much closer than it was.
Each of the players under the rules were permitted to make re-buys to carry out the experiment. After each poker playing session, the computer designers fine-tuned the computer to make adjustments in its play.
An earlier version of Libratus known as Claudico was defeated by a different group of poker professionals in early 2015.
Observers said the computer won by making bluffs and over-betting pots along with making value bets that confused its human foes.
All I can say after hearing about this story is that I am glad all my opponents are human.
Author: Geno Lawrenzi Jr.
(Geno Lawrenzi Jr. is an international journalist, magazine author and ghostwriter and poker player who lives in Phoenx, AZ. He has published 2,000 articles in 50 magazines and 125 newspapers. If you want to share a gambling story or book idea with him, send an email to glawrenzi@gmail.com ).
Your feedback
Please enter your comment.
Your comment is added.
