Society of AI

Sep 29, 2020

4 min read

Do You Know — ALPHA- GO, By Mind Technologies?

Let’s Look what is this….

AI that became Checkers Champions a few years ago!!!!

Boardgame Go is a traditional game played by Korea, China, and Japan. This game is a brain game like Chess but it is much more concrete when it comes to the use of cognitive ability and heuristic skill sets. More thinking and patience is required while playing this game. Unique game is surely regarded as being played by the genius.

It was a milestone in the field of AI when the computer Deep Blue an IBM’s computer started playing a board game Go. This revolution was seen in 1997 when Garry Kasparov a grandmaster of Chess was defeated by the computer machine. Leaving a mark that humans can be defeated or a computer has a better faculty. It opened opportunities for all to do research and indeed a lot was done. When Artificial Intelligence is being used in developing computer programming to playing games as well and winning it too.

Alpha Go is developed by Mind Technologies, it utilizes a Monte Carlo Tree exploration algorithm to find the moves based on the previously recorded information memorized by machine learning particularly by an artificial neural network. This helps in creating a neural network which is trained to predict Alpha Go. Deep mind company came with lots of development in the AlphaGo upgrading it into AlphaGo Master and AlphaGo Zero. Although there is a branching factor is involved where it limits the use of traditional methods like alpha-beta pruning and tree -traversal.

AlphaGo research project was initiated in 2014. To make it possible for a machine to play this game of Board where an immense amount of discipline is required. This game has Chinese rules and it may last for two hours or so. This needs a lot of thinking processes. In October 2015 Alpha Go Computer was programmed to beat an expert human go player without flaw. A dan professional without a handicap was able to defeat Lee Sedol in a game of Go played having up to 5 levels which lasted for two hours following the Chinese rules. The match held between 9–16 of March. The winner of the game was to be rewarded with US $the million. Lee Sedol was given $150,000 million for participating in all the games despite losing being popular for professionals in these games. And he won in the fourth level which entitled him the reward of 20,000 million US $. Alpha Go was rewarded by the Badu Association. This particular match has been featured in the documentary film AlphaGo made in 2020.

Alpha Go has already defeated a European Champion Fan Hui a 2 Dan (2/9dan). In 2017, Future Go Summit Alpha Go Master defeated no.1 ranked player Kee Jei. He didn’t want to play much as he had the apprehension his skillset will be learned through machine learning by the computer. AlphaGo was then awarded a professional 9 dan from the Chinese Weixi Association.

Netfli Alpha Go
Netflix Alpha Go

Deep mind later came with the development of “ A blank page “. A neural network was built to predict the working of Alpha Go. Alpha Go Zero was the later major development that was able to achieve about 100 victories currently it is the top player in chess. On 29 December 2016, an account holder with an Account name of “ Magister “ from South Korea and Later Changed it to Master from Fox Go. Later it was confirmed that both the users were Master and Magister is Alpha Go by the deep mind. The potential being 10 games per day it was disclosed when 59th game was found to be controlled by one of the teammates of the Deep Learn company Dr. Aja Huang. it led to the coming of Alpha Go Master in this field.

On 11 December 2017, an Alpha Go Teaching Tool was launched which had all the guidelines of playing a game with the computer device powered with Alpha Go technology which can be mastered by anyone who is might be a game enthusiast. An open environment was later created giving an open chance to play with the computer Alpha Go and win rewards.

The research on this design is ongoing and continuous which is receiving the attention of all the scholars to upgrade the tactic. Logical weaknesses are also being realized in the process making a room for improvement and

Google Deep Mind
Google Deep Mind

up-gradation for it and coming up with a new successor. Definitely, Alpha Go AI is a checker champion now honoured with so many rewards, and giving a defeat to all the great players of the board go who are worldly renowned has surely added a milestone of achievement in the field of AI.