The stakes have been officially raised in the ongoing battle for gaming supremacy between man and machine. Not content with mastering chess and Go in a series of high-profile clashes, artificial intelligence is now turning its hand to poker in a "Heads-up, No-Limit" Texas Hold'em tournament that pits professional players against an advanced "poker bot" AI to see who will fold under pressure first.

The deciding competition is being held at the Rivers Casino in Pittsburgh, where Carnegie Mellon University's Libratus computer programme is taking on four poker specialists in a marathon tournament that will see more than 120,000 hands dealt across 20 days.

The tournament began on 11 January, and while early betting odds sat firmly in camp human, a series of upset victories have seen the AI underdog triumph over its mortal opponents.

A Twitch livestream has been broadcasting the results from the event (embedded above), with thousands witnessing Libratus take a $159,288 (£132,189) lead over 23,340 hands at the time of writing and winning three out of four match-ups.

Developed by Professor Tuomas Sandholm and PhD student Noam Brown, Libratus' strategy "uses algorithms to analyze the rules of poker and set its own strategy, based on approximately 15 million core hours of computation at the Pittsburgh Supercomputing Center (PSC)."

Carnegie Mellon is also keen to stress that the algorithms governing Libratus are not only applicable to poker, citing "the AI's ability to reason when faced with incomplete or misleading information" as having "a wide range of possible applications, including business negotiation, medicine, cybersecurity, auctions and more."

The matches are active between 11am–7pm in EST (4pm-12am GMT) every day until the tournament ends. Links for the four individual match-ups can be accessed by clicking the links below:

Dong Kim vs Libratus

Jimmy Chuo vs Libratus

Daniel McAuley vs Libratus

Jason Les vs Libratus