In a new Wired article, Adam Kucharski explains why poker may be more difficult/interesting to AI researchers than either chess or Go, where all strategic information is right in front of the players:
He quotes chess master Garry Kasparov (who lost to IBM's Deep Blue computer in 1997), saying that computers play games like chess and Go "like a machine." And then writes further,
"Kasparov hoped that games such as poker would be different. You cannot win by following a fixed set of rules because some cards are hidden, and your information is imperfect. The same is true of many other situations in life, from negotiations to auctions and trading."
Kucharski reports that the latest poker-playing robots "are revealing new and innovative ways of juggling risks and making decisions with imperfect information" and "The world's top poker bots have taught themselves to bluff, feign aggression and even manipulate their opponents."
One successful poker bot from Canada that Kucharski cites (and that progressively learns "by playing billions of simulated games") is "Cepheus" (specifically for a limit version of Texas hold 'em):
And, no doubt, more are on the way.