And we’re gearing up another scintillating battle in the history of humans vs AI intelligence. Google’s DeepMind division had gathered immense applause when its AlphaGo AI managed to beat South Korean world champion Lee Sedol in the ancient board game of Go in March last year. But, Google’s uber-smart but virtual Go player is traveling to China to take on the worldwide champion of the game in May of this year.
Yes, man and machine will once again meet for a formal rematch at the company’s five-day AI symposium — Future of Go Summit in Wuzhen, China. We are calling this a rematch because AlphaGo has already managed to defeat 19-year-old Go world champion Ke Jie by logging in as mysterious player ‘Master’ on Go servers online. They will finally come face-to-face for a three-game Go match at the conference, between May 23 and May 27. Demis Hassabis, CEO and co-founder of Google DeepMind said in the blog post,
The centerpiece of the event will be a classic 1:1 match of three games between AlphaGo and the world’s number one player, Ke Jie, to push AlphaGo to (…perhaps beyond!) its limits.
Google DeepMind, in an official blog post, said it is collaborating with the China Go Association and Chinese Government to bring the top minds in AI and their AlphaGo AI software to the country next month. The tech giant has chosen one of the most serene and calm water towns in China to organize the battle of human brain versus a self-learning, coded AI software.
This Go battle will also be significant because this is the first time AlphaGo will go up against the true world champion, instead of the one who previously held the title. Lee Sodol was Korea’s champion but had lost the world champ title to Ke Jie a long time ago. Thus, AlphaGo’s win against him did surprise the masses but they rebuked Google for not going up against the true world champion. And it is now finally happening next month.
Though AlphaGo’s match against the world’s number one Go player will be the highlight, but that’s not the only aspect of this massive meet centered around artificial intelligence. Google is also introducing a variety of game formats to attract several other Go players to test their might against the AI software.
These players can compete and explore the mysteries of this 3,000-year old game together. A player can either play by pairing with an AlphaGo team member and alternating moves to win against the opponent or take it to the next level by forming a five-member team to challenge AlphaGo to a match. These five players can work together to test the software’s capabilities and adaptabilities.
Though the conference is all about celebrating the advancement of artificial intelligence over the years, AphaGo and the advancement in its capabilities will be the center of attraction for everyone. And I say this because the AI-powered Go player has definitely managed to learn a handful of new techniques by re-evaluating centuries-old inherited knowledge. This progress has been detailed in a separate blog post right here, while this platform in China will be used to test out its might against the world champ.
Speaking of playing against an AI, Zhou Ruiyang, 9 Dan Professional of Go, said:
AlphaGo’s play makes us feel free, that no move is impossible. Now everyone is trying to play in a style that hasn’t been tried before.
Further, Google DeepMind is also looking to shed light on how the development of AlphaGo has given them a chance to think out of the box and explore the possibilities of using the said machine learning algos to grapple real-life problems. We’ve already been using machine learning to research existing medical projects, enhance Google’s consumer-facing tools such as Photos, Translate, etc. and significantly reducing energy usage.
Talking about the same, Hassabis continues to add,
We will explore how AlphaGo has created new knowledge about the oldest of games, and how the technologies behind AlphaGo, machine learning, and artificial intelligence, are bringing solutions to some of the world’s greatest challenges into reach.