AI Wipes the Floor with Four Poker Pros

An adaptive artificial intelligence has utterly destroyed four professional poker players at a special 20-day event at Rivers Casino in Pittburgh. The AI, dubbed Libratus, was developed by researchers at Carnegie Mellon University to implement strategic reasoning with “imperfect information.” To test its ability, Libratus took on four of the world’s top poker pros – Dong Kim, Jimmy Chou, Daniel McAulay and Jason Les – at Heads-up (meaning two-person games) No-Limit Texas Hold’em and rinsed them for (a theoretical) $1,766,250. As a consolation for the felled human players, the quartet split a $200,000 pot, based on performance.

The team has denied the claim made by the defeated human players that they were making changes to Libratus’ programming overnight to make it shift strategies, instead explaining that the AI itself made tactical decisions on the fly based on flaws it detected in its opponents’ game.

“The event was surrounded by speculation about how Libratus was able to improve day to day during the competition,” a press release from Carnegie Mellon University reads. “It turns out it was the pros themselves who taught Libratus about its weaknesses.”

“After play ended each day, a meta-algorithm analyzed what holes the pros had identified and exploited in Libratus’ strategy,” Thomas Sandholm, co-creator of Libratus, explained. “It then prioritized the holes and algorithmically patched the top three using the supercomputer each night. This is very different than how learning has been used in the past in poker. Typically researchers develop algorithms that try to exploit the opponent’s weaknesses. In contrast, here the daily improvement is about algorithmically fixing holes in our own strategy.”

The technology behind Libratus is not exclusive to poker, and Sandholm is also working on an automated system to negotiate financial deals through his company Optimized Markets.

Image courtesy of Thomas van de Weerd.