Artificial Intelligence Could Put an End to Professional Gamers

Artificial Intelligence f

I’m old enough to remember the shock waves created by Deep Blue’s victory over reigning chess champion Gary Kasparov back in 1997: everyone knows you need a giant brain to play chess and the world’s best chess brain had been beaten by AI; it was as if that supercilious symbol of AI, Kubrick’s HAL, had arrived in town and was about to put us all in our place. Except that wasn’t the case at all, because unlike life, chess is a game of complete information, now poker, that’s a different game entirely.  

Poker is a game of incomplete information and therefore it’s much more like the problem solving situations we face in life.  It’s estimated no-limit Texas hold ‘em has the potential to generate more unique situations than the number of atoms in the universe, squared.  Poker bot software has been on the market since 2008 and the New York Times was reporting it as a problem back in March 2011 but up to now the software has been very crude, easy to spot and not very good at winning.  Many online poker rooms reference AI in their anti-cheating policy:

‘We are also committed to detecting and preventing the use of software-programs which are designed to enable artificial intelligence to play on Our Platforms’partypoker   


The challenge from AI, however, is becoming more intense.  In 2015, four of the world’s best professional poker players took on AI developed by a team at Carnegie Mellon University, they won but it wasn’t exactly a crushing defeat.  The AI they were up against was able to approximate ideal rational play but couldn’t process how cards in a player’s hand affect the likelihood of another player having a particular card combination, this resulted in bluffs which were easy to spot.  

Researchers from University College London have subsequently developed Texas hold ‘em playing AI which learns by playing games with itself, without any prior knowledge of strategies.  This deep reinforcement learning is of profound significance and is why poker has been of such interest to researchers.  When AI can play poker better than humans, when it can deal with unpredictability and incomplete information better than the human brain then it can really start to improve the lives of everyone on the planet.  The implications are of course profound: it’s not just that our transport could be safer, our medical research accelerate unimaginably, our global financial systems be taken out of human hands, but according to Alain de Botton AI could also restructure our woefully inadequate emotional intelligence, helping develop greater self-knowledge, choose better relationships, benefit from education which realises our unique talents and provide us with a personalised media input which is balanced and constructive.  So, a good thing for humanity but a problem for online poker players.  Well, possibly, because cyber defence is developing every bit as fast as cyber attack and I think it’s going to be a while yet before there’s commercially available software sophisticated enough to infiltrate online poker rooms.  I think the next time you lose at Texas hold ‘em, online, you can be pretty sure it wasn’t because a bot beat you.

Chief Editor of DroidHorizon. I own a few different devices at the minute and enjoy writing reviews and sharing what I think is cool. You can often find me playing Playstation 4 or sprawled across the couch with headphones on (maybe with a nice beer or gin & tonic in-hand)

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>


Time limit is exhausted. Please reload CAPTCHA.

This site uses Akismet to reduce spam. Learn how your comment data is processed.