Opinion written for Canadian Community News by Mike Sterling
To Comment on this article Click Here
Go Master Struggles against AI Machine Named AlphaGo
When I was young I worked in a prestigious research organization. I toiled away on getting computing machines to assist in mastering an art form called lofting. We were successful. It took a long time to understand the skills required by the computer programs.
Some of the group of researchers played a game called 'Go' over lunch. If you are not familiar with it, neither was I. I was told it made chess look trivial. It looked interesting.
It's a board game that has few rules and a lot of strategy. For the simple rules and some background to the 3000 plus year old game Click Here
Artificial Intelligence, AI, is the quest to create a machine that can mimic and surpass the limits of the human mind.
Simple games like Checkers fell to computer challange long ago. A chess playing machine called 'Big Blue' defeated the best Chess player in the world at the time, Garry Kasparov on February 10, 1996. Development for Big Blue began in 1985.
It was clear that the speed and power to go deeply into chess moves doomed humans to defeat forever more. Big Blue was retired by IBM, but AI research continued.
Later, IBM developed a machine that defeated the best Jeopardy players in the world. This shocked everyone and the AI community too.
While chess is a game that involves strategy, a player like Blue could go to any depth and man was in the minor leagues forevermore.
But, Jeopardy seemed different because nuance was a prime factor. How about this game?
This and natural language, some thought, would defeat the machine called 'Watson'. .
Indeed the initial trials brought smiles to the faces of the popular Jeopardy team and their championship players. They easily defeated Watson and IBM went back to the drawing board. Returning later, they laid waste to the Jeopardy Champions and the machine took away a $1,000,000 prize. To view a documentary on Watson Click Here
Watson has come a long way and now is IBM's leading product. It is useful in medical research and much more. I consider it the best potential breakthrough product in computing.
But, the AI community still needed more. Were they just dealing with power and speed?
How about a game that was very simple, but epitomized the need for deep strategy. Humans are supposed to excel at strategy and machines in speed and database indexing.
Google is now big on artificial intelligence and purchased a company called 'DeepMind' recently. With a program DeepMind developed called AlphaGo, it was able to defeat European Champion Fan Hui and then legendary Go Player Lee Sol.
This is a big deal for AI. A game like Go had been beyond AI capabilities, but no more. Another domino has fallen.
How close is the tipping point? Better still, what the heck is the tipping point? Is it changing?
AI as typified by Watson and OpenMind's research make us look over our shoulder. While some play around with fancy phones and pads, there is a revolution taking place that gets little publication. It's really important.
If you would like to examine related or unrelated articles, enter a key word or phrase in the search engine box below to search the Canadian Community News online database.
Click on the ads for more information
books, sports, movies ...
Wednesday, April 13, 2016