Gizmos Freeware Reviews  

Go Back   Gizmo's Freeware Forum > Debating Chamber > ChitChat

Reply
 
Thread Tools Display Modes
Old 13. Mar 2016, 06:27 AM   #1 (permalink)
Maestro di Search
 
Jojo Yee's Avatar
 
Join Date: Jul 2008
Posts: 7,835
Default Computer AI wins over human brains

In 1997, IBM-developed supercomputer Deep Blue beat world-class chess champion Garry Kasparov.

Now, Google-developed AlphaGo beat world-class Go champion Lee Se-Dol and won $1 million in an unassailable lead 3-0.

Until now, the game of Go has been considered the exclusive territory of humans as it is more complex than chess and more difficult for the AI to "think" like a brain. The first move of a game of chess offers 28 possibilities; the first move of a game of Go can involve placing the stone in one of 361 positions.

http://news.yahoo.com/computer-wins-...082146733.html
Jojo Yee is offline   Reply With Quote
Old 13. Mar 2016, 12:24 PM   #2 (permalink)
Senior Member
 
Join Date: Jul 2010
Location: New Zealand
Posts: 1,741
Default

I've not heard people say Go is more complex than chess. I do know that Go has a vastly greater number of possible games simply due to the larger size of the board. But in many respects it is simpler computationally as there is only one counter type and counters don't move.

Maybe views on this are a matter of cultural superiority: Go is popular in Asia but not where I live.

Chess is the more popular game globally so that has been the focus of AI efforts. If Go had been more popular in the West then a Go champion would probably have been beaten much earlier.
__________________
Better to light a candle ... than to curse the darkness.
Remah is offline   Reply With Quote
Old 13. Mar 2016, 06:02 PM   #3 (permalink)
Senior Member
 
Join Date: Jul 2010
Location: New Zealand
Posts: 1,741
Default

Ha ha, I stand corrected. This is a sensational achievement.

I read the Computers and Go and AlphaGo program at Wikipedia and they are very interesting. Amazing that the program advanced so much in a year.

The lack of successful optimising strategies for the computer means they had to rely on brute force approaches. The problem is that calculating exhaustive solutions exhausts the computer program because there are so many combinations even if only calculating 7 or 8 moves ahead. That is a bit of a disadvantage when a human player can glance at the board and "see" moves in a split second.

The successful strategy is based on being able to accurately estimate the probability of success which might well be the key to their success:
Quote:
AlphaGo does not attempt to maximize its points or its margin of victory, but tries to maximize its probability of winning.
__________________
Better to light a candle ... than to curse the darkness.
Remah is offline   Reply With Quote
Old 14. Mar 2016, 05:07 AM   #4 (permalink)
Maestro di Search
 
Jojo Yee's Avatar
 
Join Date: Jul 2008
Posts: 7,835
Default

Quote:
Originally Posted by Remah View Post
I read the Computers and Go and AlphaGo program at Wikipedia and they are very interesting. Amazing that the program advanced so much in a year.
Indeed, Remah, it's a breakthrough for AlphaGo to apply deep learning, instead of brute force, to beat the world's champion of Go.

According to this article, "When Deep Blue topped world chess champion Gary Kasparov in 1997, it did so with brute force... Thatís simply not possible with Go. In chess, at any given turn, there are an average 35 possible moves. With Go, there are 250. And each of those 250 has another 250, and so on. As Hassabis points out, there are more possible positions on a Go board than atoms in the universe."

To cope with this, AlphaGo achieved a great milestone by using the technique of deep learning, relying on neural networks that analyze large amounts of data in an effort to 'learn' a particular task. "Feed it enough spoken words, and it can learn to recognize what you say. Feed it enough Go moves, and it can learn to play Go."
Jojo Yee is offline   Reply With Quote
Old 14. Mar 2016, 05:13 AM   #5 (permalink)
Maestro di Search
 
Jojo Yee's Avatar
 
Join Date: Jul 2008
Posts: 7,835
Default

Now "the rise of the machines came to a halt, temporarily at least, when the champion Go player Lee Sedol beat a computer program on Sunday to prevent a whitewash after losing the first three games.

AlphaGo, developed by the Google subsidiary DeepMind, has an insurmountable lead in the series, but Sedolís win restored some human dignity."

Read more: Go humans: Lee Sedol scores first victory against supercomputer
Jojo Yee is offline   Reply With Quote
Old 14. Mar 2016, 10:19 PM   #6 (permalink)
Editor
 
George.J's Avatar
 
Join Date: Oct 2010
Posts: 1,932
Default

Quote:
AlphaGo seemed to have a problem with unexpected moves, indicating the machine lacked the ability to deal with surprises
There is also similar quote from Interstellar..... and it's the reason why AI can never match up with Humans .....ever....
__________________
If you seek for attention, do common things in life in an uncommon way!
George.J is offline   Reply With Quote
Old 14. Mar 2016, 10:56 PM   #7 (permalink)
Senior Member
 
Join Date: Jul 2010
Location: New Zealand
Posts: 1,741
Default

As Alphago calculates a probability estimate then it is quite likely that the underlying model does not accurately account for moves that are unusual, unexpected or unpredictable.
__________________
Better to light a candle ... than to curse the darkness.
Remah is offline   Reply With Quote
Old 15. Mar 2016, 07:15 AM   #8 (permalink)
Editor
 
George.J's Avatar
 
Join Date: Oct 2010
Posts: 1,932
Default

I don't think an AI can improvise and adapt to unexpected situations like a human. They don't perform risks but stick to textbook methods.
__________________
If you seek for attention, do common things in life in an uncommon way!
George.J is offline   Reply With Quote
Old 15. Mar 2016, 11:36 AM   #9 (permalink)
Maestro di Search
 
Jojo Yee's Avatar
 
Join Date: Jul 2008
Posts: 7,835
Default

Final result: 4-1, AlphaGo won $1 million and Lee Sedol $170,000 for participation and won the fourth game.

Wikipedia has an immediate update: https://en.wikipedia.org/wiki/AlphaGo_versus_Lee_Sedol

Match 5 Live on YouTube: https://www.youtube.com/watch?v=mzpW10DPHeQ
Jojo Yee is offline   Reply With Quote
Old 19. Sep 2016, 04:54 AM   #10 (permalink)
Member
 
Join Date: Sep 2016
Posts: 0
Default

human owns the power to invented computer AI and also has the power to destroy it.
lotusleaf is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



All times are GMT +1. The time now is 05:37 PM.


Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2020, vBulletin Solutions, Inc.