Draws parallels to the 1983 movie WarGames
Jan. 28, 2019
The Pentagon has hired a critically-acclaimed ‘AI robot’ for wargaming.
The robot, Libratus, made headlines in 2017 after it defeated four top poker players during a Texas Hold’em championship, a game in which, just like a war scenario, players don’t have full intel on their opponents, such as what cards the other players are holding.
Naturally the Pentagon would be interested in a system that won a poker championship using algorithms meant to crack information-imperfect situations.
“The Pentagon has bought Libratus for a cool $10 million,” reported NextBigWhat.com. “Two researchers from Carnegie Mellon University built Libratus using a technology called computational game theory.”
The news might bring to mind the 1983 movie WarGames, in which a NORAD supercomputer, named WOPR, is programmed to constantly run war simulations and learn over time:
- A d v e r t i s e m e n t
One of Libratus’ designers, Professor Tuomas Sandholm, founded a startup called Strategy Robot to modify the technology for government purposes, according to media reports.
“To think of Libratus as just a poker-playing champ is to sorely underestimate it,” reported Engadget in 2017. “Instead, Sandholm says, it’s a more general set of algorithms meant to tackle any information-imperfect situation.”
As we previously reported, there’s three major factors behind current AI development: the progression toward quantum computing, the advancement of “deep learning” in which computers are taught to learn on their own, and the availability of Big Data.
It appears that Libratus used “deep learning” techniques to win the poker match.
The development of quantum computing, which is known publicly to still be in its infancy, would revolutionize robotics because quantum computers utilize the ability of subatomic…