Computer geeks in the
US have finally created a robot that can play poker, and like the song says, he knows when to hold them, knows when to fold them, and knows when to leave them and run.
Carnegie Mellon University computer scientists developed a program that they have given the name, smart learning, because it can play a winning hand of Texas Hold Em.
A professor and a graduate student are the creators of this game theory program that will be placed in competition in the new Computer Poker Competition which will be run by the American Association for Artificial Intelligence. The tournament will take place in Boston over the 16 to the 20 of July.
GS1, as the robot is called, cannot beat the best of the human players just yet. However, since it has beaten the two leading computer poker robots in heads up play this year, it looks good for GS1 in the poker tournament strictly for robots to place well in the tourney.
Most programs for robots playing poker are based upon the expertise and know how of human poker players. But GS1, is a formulation of skill and strategy, plus theory based analysis of poker rules and good circuitry.
The two genius men from the Carnegie Mellon University, Sandholm and Gilpin, have since the announcement of GS1, have improved upon their program and created a more advanced prototype, GS2, which may have a chance to play against the original GS1.
