The short answer is: yes, very much. Firstly, it is a kind of a benchmark as to how far artificial intelligence is along. Go is a very difficult game, and a game of perfect information – there is no luck involved. Secondly, DeepMind have pushed some of the boundaries and techniques in intelligent decision making and optimization.
But third, and most intriguing to me, is that I believe in the future, we may solve tough real world problems by encoding them as game positions.
Do you know about The Treachery of Images?
 |
| This is not a pipe. |
This is exactly the same: there is a fundamental difference between Go and Life.
In Go you know all the rules, and play in an isolated environment.
In Life, you don't: none of the above preconditions apply!
This is what the total history of science is about: realizing that we were wrong, find new rules (a bit deeper understanding of the fundamental laws of nature), and step forward as a civilization by using them. And then realize where we were wrong, and do it again.
So showing that now we have enough performance to run a massive algorithm that finds, evaluates and optimizes billions of actions among known rules better than a human player is, well, kinda nice, and yes, it is important because now you can calculate the trajectory of a rocket or satellite with more precision than legions of human computers with pencil and paper.
But to solve (and create...) problems, it is still and always us. Human beings.
However, to solve the problems that WE create, we must return to the level we named our kind about: Homo "Sapiens". Wise, not intelligent. And here a game playing toaster does not help much, only distracts us from our own tasks and responsibility.
An important tool, but definitely not the key to the future.
