Go - A Game to Learn

Often games are part of childhood memories and each of us may have played several board games that were not only intriguing but hold a deeper meaning to existence. Games are a microcosm of real-life strategy and problem-solving skills. 

Many of these children grow to be the likes of Bobby Fischer, Vishwanathan Anand, Garry Kasparov, or Lee Sedol, These are some of the great human minds or Tendulkar of board games. 
Games lead to an interesting evolution in modern-day computing and other technological developments such as deep learning, in simple terms machine learning or artificial intelligence (AI). Refer to my last blog on AI and there is a glimpse of challenges thrown at the human race by them. 
Who will be the Master?

It can mimic the brain functioning of human beings in detecting objects, recognizing speech, Translate language and make decisions, Deep Learning AI is able to learn without human supervision and use data to draw interference. AI Guru Kai-Fu-Lee predicts nearly 50% of the jobs will be lost to AI in the next 15 years. 

Let me indulge you with some interesting events that took place in the last decade where it all began,  Deep Blue Versus Gary Kasparov in 1996 when a Supercomputer designed by IBM played a six-game chess match. The first match was played in Philadelphia and won by Gary but lost the second match in 1997 in NY which was won by Deep Blue, the other three games were drawn with mutual consent but Deep Blue was victorious in the last game. 

It brings me to an interesting game called “Go” most simple to learn but impossible to master.
  “So Go is probably the most complex game ever devised by man. It has 10^170 possible board configurations, which is more than the numbers of atoms in the universe,”
said study author and AlphaGo co-developer Demis Hassabis of Google DeepMind. The team's goal was to beat the best human players, not just mimic them.

Go has a different board configuration unlike chess, and they have boards in  9, 13, and 19 board configurations, played with black and white stones. 

Objective: The winner is the one who gains maximum territory by capturing the opponent's stones.

Eighteen times world champion Lee Sedol of South Korea is a celebrated national hero of Seoul and ranks 9 Dan (refer to the ranking chart given below) professional in Go. He is a Sachin Tendulkar of Go and celebrated in Korea, Japan, and China.

Google designed an AI Alpha Go program and decided to field against the best players, it uses an algorithm which is a combination of Machine learning and Tree Search technique and training from real-world recorded games. Systems' entire Neural network was bootstrapped with human game-play.

The match was arranged AlphaGo versus Lee Sedol between 9-15 March 2016, Lee was confident of beating its opponent but lost his first match but defeated AlphaGo in his fourth game but lost with 5-1, Match prize money was $1 Million and Lee received $170000 ($20k for winning 1 match and $150K for participation)

Google donated the prize money to several charities.

 GO ranking system is tabulated from the lowest to highest ranks:

Rank Type	Range	Stage
Double-digit kyu (級,급) (geup in Korean)	30–20k	Beginner
Double-digit kyu (abbreviated: DDK)	19–10k	Casual player
Single-digit kyu (abbreviated: SDK)	9–1k	Intermediate amateur
Amateur dan (段,단)	1–7d	Advanced amateur
Professional dan (段,단)	1–9p	Professional Player

(There is also an amateur title of 8-dan and a professional title of 10-dan, but these are not the same as ranks.)

The above episode had several outcomes some of them are:

1. South Korean government decided to Invest $863 million (1 trillion won) in artificial intelligence (AI) research over the next five years in 2016.

2. Increased dependency on Machines in decision making human factor redundant.

What is your take on the blog, please share your views?