Look Out, This Poker-Playing Computer Is Unbeatable : All Tech Considered Forget chess. Poker may be even harder for a computer. But a software program has now "solved" a variant of Texas Hold'em, the bot's creators say. And nothing can keep it from winning.
NPR logo

Look Out, This Poker-Playing Computer Is Unbeatable

  • Download
  • <iframe src="https://www.npr.org/player/embed/375736513/375923274" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript
Look Out, This Poker-Playing Computer Is Unbeatable

Look Out, This Poker-Playing Computer Is Unbeatable

  • Download
  • <iframe src="https://www.npr.org/player/embed/375736513/375923274" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

ROBERT SIEGEL, HOST:

Today in the journal Science, researchers introduced a computer program that can outwit any human on the planet at poker. To help explain how this gambling robot beats the odds, NPR's Geoff Brumfiel goes back to last year's World Series of Poker.

(SOUNDBITE OF 2014 WORLD SERIES OF POKER)

UNIDENTIFIED ANNOUNCER: And there it is, the pocket nines for Jacobson.

GEOFF BRUMFIEL, BYLINE: The final was broadcast on ESPN. This particular type of poker is called Texas Hold'em. Don't worry if you don't know the rules, players have cards, they bet, then they show their hands. And one person takes the pot.

(SOUNDBITE OF 2014 WORLD SERIES OF POKER)

UNIDENTIFIED ANNOUNCER: Martin Jacobson has his breakthrough win. He has climbed the mountain and is poker's 2014 world champion.

BRUMFIEL: Poker isn't just a game of chance. It's a mixture of luck, statistics and psychology. You need to know the odds, know whether your opponent is bluffing. You've got to know when to hold 'em, (singing) know when to fold 'em, know when to walk away. Enough. Michael Bowling is a computer scientist at the University of Alberta in Canada who invented the new program. He's played poker in Vegas once.

MICHAEL BOWLING: I actually did phenomenally well out of pure luck and so I'm just going to never gamble again and be up on Vegas.

BRUMFIEL: Bowling loves thinking about all the strategies for playing poker.

BOWLING: Well, there are 10 to the 14, which is 10 trillion different decision points in the game. And so the actual number of ways of playing is an astronomically large number.

BRUMFIEL: And it turns out, one of these trillions of strategies can beat all the others. It doesn't matter if your opponent is a master bluffer or blundering amateur; a computer with that strategy can eventually win. It took Bowling years to find it. His team had to crunch the strategies with supercomputers. But eventually, they came up with the perfect strategy; well, almost perfect.

BOWLING: It is just a tiny bit off, and that tiny bit is so small that even if you played a lifetime -12 hours a day, 200 hands an hour for 70 years - you still wouldn't be able to tell it apart from actually having played a perfect game.

BRUMFIEL: It does have limitations. It can only play one opponent at a time. And the version of poker it wins at that isn't exactly like what's on TV. But that doesn't matter to Bowling. Remember, he's not a gambler.

He hopes the program he's developed can make computers generally better at making decisions when faced with uncertainty. Still, this left me wondering - are there any games for humans to win? I called Oren Etzioni, the head of the Allen Institute for Artificial Intelligence and asked him, maybe Connect Four?

OREN ETZIONI: The computer will never lose in Connect Four.

BRUMFIEL: Checkers.

ETZIONI: The computer will beat you.

BRUMFIEL: Backgammon.

ETZIONI: The computer is playing at the level of the world champion.

BRUMFIEL: Parcheesi?

ETZIONI: I don't know Parcheesi. It sounds kind of cheesy. (Laughter).

BRUMFIEL: I don't think anyone plays Parcheesi anymore. Maybe I can win on that ground.

ETZIONI: Yeah.

BRUMFIEL: Actually, Etzioni says humans still have an edge. Chess computers, poker computers, Parcheesi computers, if they exist, they're all completely different.

ETZIONI: It's not the same program that's doing this. It's very, very different computer software.

BRUMFIEL: A single human brain, on the other hand, can play all those games even if it loses. Geoff Brumfiel, NPR News.

(SOUNDBITE OF SONG, "THE GAMBLER")

KENNY ROGERS: (Singing) You've got to know when to hold 'em, know when to fold 'em, know when to walk away and know when to run. You never count your money when you're sitting at the table. There'll be time enough for counting...

Copyright © 2015 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.