This AI-Poker bot won $1.7 million defeating the world’s top pros

Over the past month, an AI poker bot called Libratus played thousands of games of heads-up, no-limit Texas hold'em against top professional players at Rivers Casino in Pittsburgh. What's more, it beat them all.

The Game

The Brains versus Artificial Intelligence competition saw four human players – Dong Kim, Jason Les, Jimmy Chou and Daniel McAulay – spend 10 hours each day stationed at computer screens – playing against a piece of software at no-limit Texas Hold'em. Libratus defeated them all, winning more than $1.7 million in chips.

Machines have already become smart enough to beat humans at other games, for example, chess and Go, yet poker is more troublesome because it's a game with imperfect information. With chess and Go, each player can see the entire board, however with poker, players don't get the opportunity to see each other's hands.

Furthermore, the AI is required to bluff and effectively interpret misleading information keeping in mind the end goal to win.

Regardless of the pure ability of the humans and the AI, it seems clear that the pros will be less compelling as the tournament goes on. Ten hours of poker a day for 20 days straight against an emotionless computer was exhausting and demoralizing. And keeping in mind that the humans sleep at night, Libratus takes the supercomputer powering its in-game decision making and applies it to refining its overall strategy.

For the AI, however, this is just a beginning. Having demonstrated that it's conceivable to beat the professionals, there is now a clear context to bite on: multi-player no-limit Texas hold'em. In any case, the game theory used as a part of the present software falls down when there's more than one opponent, and it's not clear what method to use.