Watching the live Google Deepmind Go Challenge: Lee Sedol vs AlphaGo

It’s 1h30 am, on a Tuesday night. Yet I’m up, in front of my computer, like a complete zombie. Nope, it’s not another gaming night, nor a watching someone else game, à la Piediepie.

Nope, I’m watching a computer play against a human, on a very simple game, that shows incredible computational challenges. I’m watching a historic moment.

The commentary is excellent. Every move is shown several variations. Holding territory, and wagging war, is all a choice. Discard this to get that. Exchange that area for another position. Sacrifice this land to capture this one.

Bite like a tiger. Set a net. These are the basic fundamental ideas that humans have thought since millennia. The game, in which stones are set, but never moved, can be represented as an abstraction of so many difficulties of choice that us humans are faced with.

How do you play the long game? Every move you makes accumulates. Energy spent can never be salvaged.

Tonight, Google sets it’s AI, “alphago”, against Lee Sedol, a 9-dan master. It’s the first of up to 5 matches, but really, this first one is the most important. If alphago wins, and the next 4, then who knows how better this AI is than us for these simple (to define), yet hard to win (to solve), problems than us.

Yet, if it’s a mix, then maybe we still have that ability to beat “the machine”.

Even better, if the AI initially wins, and later on loses, it might show that us humans still have an adaptive edge, and that we won’t be obsolete so soon.