Best of The Daily Galaxy –July 6-13
Gigantic Species of Leopard-Eating Chimps Rule the Congo Jungle

Man vs Machine: Computers have beat man at every game but one—the most ancient of them all

Deep_blue Just over ten years ago, the IBM supercomputer program Deep Blue beat world chess champion Garry Kasparov—the greatest chess mind alive. That moment marked a turning point in the relationship between man and machine.

Now the computer is dominant in nearly every board or card game devised by man. Computers have superior command of chess, draughts, Othello (Reversi), Scrabble, three-dimensional noughts and crosses, Monopoly, and even at bridge and poker.

Why? Because the computer has a near flawless strategy that man is not intellectually capable of. It runs the board position through a databank and chooses the most logical next move…every time.

The inevitable progress of computer intelligence has long been a principal theme of science fiction. Indeed, many experts in the field of artificial intelligence say that AI will eventually evolve far beyond all human physical and intellectual capacities.

Yet there is one game at which the computer is still no match for humans—the ancient Chinese board game Go. It is the oldest game in the world, and it is perhaps the only game at which man remains the undisputed champion.

Go is believed to have originated long before there was writing to record it. According to legend, it was invented by an emperor who wished to teach his foolish son the virtues of balance and patience.

The game involves a simple grid board of 19 lines and two players, one with white stones and the other with black. The object is to stake out a larger territory by tactically placing the stones and surrounding the opponent's forces.

So why are human’s still master of this particular game? Because Go requires more than brute computational power, which is how Deep Blue and similar programs excel. The qualities that mark out the master Go player are the hallmarks of human intelligence: adaptation, intuition, and the ability to plan for the future. In order for a computer to win, a program has to be developed that can think more like a person.

Go has become a cryptic symbol of unlocking the secret of artificial intelligence. If a machine can figure out how to dominate at Go, it is believed that mankind would be very close to replicating human thought. However in spite of over a million dollars in prize money up for grabs, programmers have not yet been successful.

It has been calculated that there are more distinct games of Go than atoms in the known universe. A move early in the game can affect the passage of play hundreds of moves later. The vastness of the possibilities offers a wide range for individuality, and an intuitive awareness of the other players feelings and thoughts.

According to some experts, this very generation will live to know intuitive, “feeling” computers. When that day comes, we won’t just lose at Go, we’ll lose our last claim of superiority.

Posted by Rebecca Sato

Link 1

Link 2

Link 3


I play tic-tac-toe against a computer and it never beats me. Also I rock at mini-putt!!

Rebecca: You may be aware that, prior to the defeat of a human by a machine in chess, the very same argument that you make here ("Go requires...adaptation, intuition, and the ability to plan for the future") is the one that was routinely made when defending human chess players against machines.

At the present rate, researchers have estimated that machines will beat humans at Go within the next decade. The trend line for this achievement is fairly well-defined. I would be willing to make a wager on it.

If you're interested in some research that relates the relative complexity of these tasks, please see,345152.shtml
But I guess that was just a stroke of luck...

Hi Rebecca - may we reproduce this article in our new course about the environment please?
Thanks very much

There seems to be a big difference in possibilities between Chess and Go. While a real AI would have to use "adaptation, intuition, and the ability to plan for the future" chess AIs afaik just have big charts of all possible outcomes and picks the best one for the current board state. Chess has 10^123 possible games. Go has a number we can't compute, never mind being able to have a charts big enough for a computer to play go the way chess is played. That isn't to say that it won't be possible, giving a limit to computational power and memory size is pretty stupid at this point. All I'm saying is that chess it is alot simpler, and that it would probably be easier to make an AI than it would to have enough storage and power to brute force Go.

Really interest, thanks Rebecca. I am a huge fan of the game Go and yet didnt know about its history! I found the online version here

As far as I know, an average human player still beats the best computer program in most varieties of poker, the exceptions being Pot-Limit, heads-up games. If you have any reference to the contrary I'd love to hear about it.

Theres a lot of poor information here. Chess is not enumerated completely and then a "best" move picked. This is beyond the capabilities of current machines. Instead machines will investigate possible paths of the game tree (called MiniMax) and then evaluate the position at a certain depth and choose the move which leads to the best outcome.

Go is likely evaluated in the same way and may just have more options to investigate and therefore not be able to investigate to the same depth. This is not an impossibility but an eventuality.

Deep Blue cheated, as the IBM programmers worked on the program during the game, changing and correcting it.

Then, it was totally disassembled and destroyed immediately after that one game!

Bad form, old boy! You cheated, and then hid the evidence, and, though
challenged, did not respond!

Dare I say that IBM only sought notoriety, fame, and media hype? And, that they were crass enough to cheat and steal the one victory out of three losses?!

I don't know wether existe a better than deep blue as software and free. please my email:

Try winning at Double-Cranko

What about Shogi? Japanese version of chess... kind of. Highest level of complexity known to man.

I love playing Go to bad there is no one in the state of arkansas who also knows how to play it.

Hey what about the rumors I heard that Eugenio Torre beat that computer?

You do know that there were multiple grand masters in the back telling the computer what to do. The computer alone will never beat a human. It's just not smart enough to play positional, it only plays tactics.

Aaaaaaand it's done. AI has beaten humans at Go. Now what?

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Your Information

(Name is required. Email address will not be displayed with the comment.)