AI Masters Poker. Is Cybersecurity Next?

An artificial intelligence program developed by researchers at Carnegie Mellon and a Facebook AI Lab has more than held its own against some of the best poker players in the world. Moreover, it did so without recourse to the “tell,” the small physical signs in response to what they are holding that even good poker players sometimes let slip. What this program does is “Pure numbers and percentages. It is solving the game itself,” said Darren Elias, a brash poker player who has won a record four world championships but who recently fell before the automated program, which has been dubbed Pluribus. It was developed for $150, without a  lot of computer power and over a little more than a week, during which time it analyzed the results of trillions of hands that it played against itself. An expert in this kind of AI says a similar process could be applied to other situations that also involve what’s been called “hidden information,” notably to the cat-and-mouse game between hackers and cybersecurity operatives.