New Delhi: From defeating human professionals at the world’s most popular form of poker to solving the Rubik’s Cube faster than any human, artificial intelligence (AI) programmes have been overcoming unique challenges this week.
An AI named Pluribus, developed by researchers from the Carnegie Mellon University in the US, in collaboration with Facebook AI, defeated leading professionals in six-player no-limit Texas hold’em poker.
The programme defeated poker professional Darren Elias, who holds the record for most World Poker Tour titles with four, and Chris Ferguson, winner of six World Series of Poker events. Each pro separately played 5,000 hands of poker against five copies of Pluribus.
In another experiment involving 13 pros, all of whom have won more than US $1 million playing poker, Pluribus played five pros at a time for a total of 10,000 hands and again emerged victorious.
‘Milestone in AI & game theory
“Pluribus achieved superhuman performance at multi-player poker, which is a recognised milestone in artificial intelligence and in game theory that has been open for decades,” said Tuomas Sandholm, a professor at Carnegie who developed Pluribus with Noam Brown, a research scientist at Facebook AI.
“Thus far, superhuman AI milestones in strategic reasoning have been limited to two-party competition. The ability to beat five other players in such a complicated game opens up new opportunities to use AI to solve a wide variety of real-world problems,” Sandholm said in a statement.
Brown added: “Playing a six-player game rather than head-to-head requires fundamental changes in how the AI develops its playing strategy. We’re elated with its performance and believe some of Pluribus’ playing strategies might even change the way pros play the game.”
Michael Gagliano, who has earned nearly US $2 million in poker career earnings, also competed against Pluribus, and said: “It was incredibly fascinating getting to play against the poker bot and seeing some of the strategies it chose.”
Also read: Can Artificial Intelligence help us with our voting decisions?
Solving Rubik’s Cube
Researchers at the University of California, Irvine (UCI), created an algorithm called DeepCubeA, which can effortlessly solve the Rubik’s Cube, finding the shortest and quickest way to do so 60 per cent of the time.
“The solution to the Rubik’s Cube involves more symbolic, mathematical and abstract thinking, so a deep learning machine that can crack such a puzzle is getting closer to becoming a system that can think, reason, plan and make decisions,” said Pierre Baldi, a professor at UCI.
“Our AI takes about 20 moves, most of the time solving it in the minimum number of steps,” Baldi said.
A bigger challenge than chess
Games such as chess and Go have long served as milestones for AI research. In 2017, Google’s DeepMind AlphaGo artificial intelligence defeated the world’s number one Go player Ke Jie.
The same AI, when repurposed to play chess, went on to beat Garry Kasparov, the multiple-time world champion and longest-reigning number one in history (1986-2005).
In both those games, all the players know the status of the playing board and all of the pieces.
However, poker poses a bigger challenge because players do not know which cards are in play, and opponents can bluff.
That makes it both a tougher AI challenge and more relevant to many real-world problems involving multiple parties and missing information.
Ultimate goal
AI systems are slowly becoming part of our everyday lives, through chatbots, recommendation engines, and apps such as Apple’s Siri and Amazon’s Alexa.
The ultimate goal of projects like Pluribus and DeepCubeA is to build the next generation of AI systems, to address societal problems in a cost-effective manner.
Efficient AI may help treat cancer and other diseases, improve cybersecurity, manage manpower and assets more efficiently, and even pave the way for social robots to assist the differently-abled people or the elderly.
Also read: Artificial intelligence can take your job, so political leaders need to start doing theirs