Tuesday, January 17, 2017

An Artificial Intelligence is Dominating Texas Hold‘em Poker


~hehe the GREAT battle    ...human vs  inhuman ?  LOL y don't we just ask ole victor franky ....hows that work~in out 4 ya  HUH  hows that go~in     umm uh oh yea "his" monster is STILL chas~in his ass all fucking over the n. pole Oops

It’s all over. Let’s just go ahead and get used to it. While science fiction tales have foretold the oncoming artificial intelligence uprising for a while now, it used to be easy to write these off as fanciful yarns spun by the Luddite imagination. With every new AI advancement, however, it seems we might be already witnessing the end of the domination of human intelligence on Earth. Artificial intelligence systems already handily defeat puny human minds at Jeopardy, the ancient Chinese game Go, earning kills in deathmatch video games, and even picking winning horses in the Kentucky Derby. Hell, they’re even writing pop songs that are a lot better than some of the mindless drivel on the radio these days. Darned kids and their Walkmans…
What's next, AI writing daily news blogs? Wait, no, not that. Anything but that.
What’s next, AI writing daily news blogs? Wait, no, not that. Anything but that.
At any rate, artificial intelligence’s terrifying unrelenting march towards human enslavement continues with the most recent announcement that AI can now out perform even the best human players at No-Limit Texas Hold’em Poker. The system, called Libratus, was designed by computer scientists from Carnegie Mellon University. It is currently squaring off against poker pros in a twenty-day tournament at Rivers Casino in Pittsburgh. The tournament, called “Brains Vs. Artificial Intelligence: Upping the Ante,” follows a similar tournament in 2015 in which humans came out on top.
Human players competing against the previous CMU poker bot in 2015.
Human players competing against the previous CMU poker bot in 2015.
Poker is especially difficult for computer systems to play due to the complexity of both the game mechanics and human behavior. In a hand of poker, there are upwards of 10160 possible game scenarios – more than the total number of atoms in the known universe. Couple that with the human unpredictability of bluffing and betting, and you’ve got an incredibly complex game. Libratus’ designers are secretive about the system’s actual programming, but claim that the system is designed to determine every possible scenario for each turn of play and choose the best course of action.
Poker pro Jason Les squares of against Libratus as Dr. Thomas Sandholm, the AI's designer, watches on
Poker pro Jason Les squares of against Libratus as Dr. Thomas Sandholm, the AI’s designer, watches on.
Already, Libratus has defeated some of the best poker players in the game today. When some players complained about the system’s painfully slow rate of play, its creator, CMU professor Thomas Sandholm, claims Libratus’ relaxed pace is due to its intelligence:
It’s thinking. It’s thinking much faster, but it’s also thinking much more.
According to CMU, the system isn’t just adept at playing poker. Its designers claim Libratus is capable of responding to any scenario featuring “incomplete and misleading information,” such as military strategy, business negotiations, cybersecurity systems design, and even creating medical treatments. Yep, sounds like Skynet to me. If you need me, I’ll be in my doomsday bunker learning how to milk cockroaches.

No comments:

Post a Comment