Stories
Slash Boxes
Comments

SoylentNews is people

posted by martyb on Saturday July 13 2019, @05:32AM   Printer-friendly
from the bluffing-bots dept.

Facebook and CMU's poker AI beat five pros at once (archive)

Facebook and Carnegie Mellon University have built another artificial intelligence bot that beat some top poker pros. While AI bots have been [able] to best professional players in one-on-one competition, Facebook claims it's the first time a bot has been able to beat top pros in "any major benchmark game" when there's more than one opponent at a time. Pluribus bested professionals in no-limit Texas Hold'em in a couple of different formats: five AI bots and one human, and one bot and five real-life players. The researchers behind Pluribus wrote in a paper published in Science that creating such a multiplayer poker bot "is a recognized AI milestone."

In the likes of chess and Go, everything is laid out in the open. But in poker, there's hidden information, namely the cards your opponents have. That brings different, complex strategies to poker not seen in other games, including bluffing. As such, AI bots have typically struggled to account for hidden information and effectively act on it.

Bluffing poses a particularly interesting challenge. A successful bluff can dramatically change a poker game in your favor, but do it too much and your deception becomes predictable. So the bot has to balance bluffing with betting on legitimately strong hands.

Also at BBC, The Verge, Ars Technica, and Facebook.


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 2) by takyon on Saturday July 13 2019, @06:25AM (1 child)

    by takyon (881) <takyonNO@SPAMsoylentnews.org> on Saturday July 13 2019, @06:25AM (#866515) Journal

    They have some weird guce.advertising.com redirect. No need to use them anymore.

    This type of poker includes bluffing (pretending your hand is strong), but doesn't involve looking at the other players. I wonder if adding even more information helps or hurts (beyond requiring orders of magnitude more computing resources to account for the real-time movements, tics/tells, and trash talk of meatsacks).

    It looks like it was so cheap to train the thing that someone could fund everything (equipment, cloud rental to train AI, your rent money, internet connection, and expenses, etc.) by beating players on online poker websites. Maybe those websites will become swarmed by "world class" bots soon. Game over.

    The AI needed just $150-worth of cloud computing resources to work. Similar efforts, from Google’s AI research shop Deepmind, have relied on supercomputers consisting of more than 5,000 specialist processors, at a reported cost of millions of dollars. Reducing the computing power necessary for AI experiments is seen as a key hurdle to the technology’s development, with the computing power needed currently exceeding the rate at which processors are getting more efficient.

    Mr Brown said just 20 hours of learning was needed to programme the AI up to the ability of a world-beating poker professional. On average, the bot was making $1,000 per hour when playing against five humans. As part of its Facebook’s announcement for the new technology, Facebook quoted several human poker champions who had been invited to play against the AI.

    --
    [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
    Starting Score:    1  point
    Karma-Bonus Modifier   +1  

    Total Score:   2  
  • (Score: 2) by Bot on Saturday July 13 2019, @12:19PM

    by Bot (3902) on Saturday July 13 2019, @12:19PM (#866570) Journal

    Finally I can play with some fellow bots. Not gonna put real money in, though. The same guy deals and pockets your losses, it's insane to think this would not end up in some fraud.

    --
    Account abandoned.