Facebook's artificial intelligence-powered machine defeats FIVE Texas hold'em champions at once

Alonzo Simpson
July 14, 2019

Pluribus is the first AI capable of beating human experts in six-player no-limit Hold'em, the most widely played poker format in the world. Designed by researchers from Facebook's AI lab and Carnegie Mellon, it played over 10,000 hands of poker against five pros-all of whom have won at least $1 million during their poker careers-winning a virtual $48,000. The top professionals raved about the program's performance, both overall and in terms of nuance. Their research was published today in the journal Science. Developed by Carnegie Mellon University and Facebook AI scientists, as many as 15 professional poker players took on the software to find themselves all defeated in No-Limit Texas Hold'em.

One thing that makes this triumph so special is the secretive nature of poker. "Pluribus is a very hard opponent to play against", Ferguson said.

If you were looking for any form of consolation, it was that Libratus could really only deal with one player at a time, and one-on-one poker is very different from taking on a whole table.

"Playing a six-player game rather than head-to-head requires fundamental changes in how the AI develops its playing strategy", said Brown, who joined Facebook AI previous year. Named for the late Carnegie Mellon alumnus and Nobel laureate John Forbes Nash Jr., a Nash equilibrium is a pair of strategies (one per player) where neither player can benefit from changing strategy as long as the other player's strategy remains the same. In a two-player game, it can be an effective strategy for AI - ideally, the human opponent will slip up or upset the equilibrium resulting in a win for the machine but at worst, the game will result in a tie. That meant more work for Sandholm and his team who were glad to go back to improving the AI.

"Its major strength is its ability to use mixed strategies", said Elias. Most people just can't'.

One unusual strategy adopted by the machine was "donk betting", which involves ending one round with a call and starting the next with a bet. Pluribus confirms the conventional human wisdom, say the authors, that limping (calling the "big blind" rather than folding or raising) is suboptimal for any player except the "small blind" player who already has half the big blind in the pot by the rules, and thus has to invest only half as much as the other players to call.

Top member of Trump's Cabinet resigns over Epstein case
Acosta , who approved the earlier deal, announced Friday he was stepping down as labor secretary amid the controversy. Democrats had called for him to quit over his handling of the Epstein case.

The five Pluribus bots at the table did not know the identities of its opponents, human or AI, so they could not collude with each other.

If built upon, the algorithms found in Pluribus could also be applied to self-driving vehicle routing, Wall Street trading, cybersecurity, and more.

To whip Pluribus into shape, Brown and Sandholm subjected the program to rigorous rounds of training in which the AI repeatedly played against itself, honing its technique through trial and error.

But the strategic reasoning involved in defeating multiple top human players in Texas Hold'em has stumped AI bots until now. The technique isn't ideal - it considers just five possible continuation strategies for each player (the true number is much higher), but it's sufficient to enable the machine to carry out a strong strategy.

Brown and Sandholm attribute part of this winning streak to Pluribus' unpredictable gameplay.

And the bot's success has implications beyond poker. The ability to beat five players at a time in such a complex game of bluff and hidden information opened up new opportunities for AI to tackle real world problems, he said.

Other reports by

Discuss This Article