Earlier this year, Carnegie Mellon University’s Libratus defeated four of the world’s best heads up players in the second installment of the Brains vs. AI Challenge. The humans lost $800,000 over the course of the 20-day contest and this month, humanity booked another loss.

DeepStack, created by researchers at the University of Alberta in Canada, played 3,000-hand matches against 31 players from 17 different countries over the course of a four-week period. DeepStack defeated each of the 11 players who finished their matches, with only one of those wins falling outside the margin for statistical significance.

The University of Alberta programmed DeepStack to bridge the gap between approaches used for games of perfect information, like chess, with those used for imperfect information games, like poker, by using “intuition” honed through deep learning to reassess its strategy with each decision. DeepStack extends the ability to think about each situation during play by using a technique that researchers call “continual re-solving”.

“Instead of solving one big poker game, it solves millions of these little poker games, each one helping the system to refine its intuition of how the game of poker works.” said Michael Bowling, a professor at the University.

“Thinking about each situation as it arises is important for complex problems like heads-up No Limit Hold’em because there are vastly more unique situations than there are atoms in the universe.” Bowling added.

Twice in as many months, two groups of humans have lost matches against artificial intelligence. The end is near, congratulations to K-33O, 2025 World Series of Poker Main Event champion.