Google hooked us on the geekiest game you've never heard of

آخرین مطالب

امکانات وب

Google hooked us on the geekiest game you've never heard of - CNET

Deep in the action of the Google DeepMind Challenge, pitting Artificial Intelligence against human intellect.
Screenshot by Claire Reilly/CNET

The jocks of the world thought the End of Days would come with explosions, bravado and full-camouflage utility belts. But the rise of the machines is currently playing out in the Four Seasons Hotel in Seoul, and it's being led by possibly two of the nerdiest groups known to man.

It's Match 1 of the Google DeepMind Challenge: AI versus human, a slow-paced, four-hour battle of strategy, complex algorithms and black and white tiles. We know what you're thinking: Time to clear your schedule and start heating the snacks.

If you haven't heard of Go, you're not alone. Largely played in China, Korea and Japan, the ancient board game requires two players taking alteate tus to place black and white tiles on a grid, jockeying to occupy more of the board than their opponent.

Think of it like chess over a bigger monochromatic battlefield. Of course, the Go players don't think of it like that, as this is the game for people who think chess is too easy (and probably too newfangled).

Designed by Google DeepMind, the artificial intelligence arm of everyone's favourite (don't be) evil overlords, AlphaGo has been programmed to clear one of the longest-standing hurdles in AI -- winning at board games. Go's 19x19 grid offers nearly uncountable variations of moves and branching outcomes, ruling out typical AI brute-force logic.

AlphaGo last played in October, beating pro player Fan Hui. The machine has had 5 months to improve its algorithms and, in tu, its strategy, by playing game after game of Go to add to its database of moves. (Don't machines get to have all the fun?)