Java-Gaming.org Hi !
Featured games (91)
games approved by the League of Dukes
Games in Showcase (804)
Games in Android Showcase (239)
games submitted by our members
Games in WIP (868)
games currently in development
News: Read the Java Gaming Resources, or peek at the official Java tutorials
 
    Home     Help   Search   Login   Register   
Pages: [1]
  ignore  |  Print  
  Go has been cracked!  (Read 15128 times)
0 Members and 1 Guest are viewing this topic.
Offline ags1

JGO Kernel


Medals: 367
Projects: 7


Make code not war!


« Posted 2016-02-08 22:33:03 »

I'm surprised the topic hasn't come up on the forum yet:

http://www.wired.com/2016/01/in-a-huge-breakthrough-googles-ai-beats-a-top-player-at-the-game-of-go/

It's interesting that even the best Go players can't explain exactly why a certain move is good, and now we have built a Go-playing computer, and we can't say precisely (or even vaguely) how it chooses the moves it chooses Smiley

Who knows if Deep Mind has not decided the most efficient way of winning at Go is to subjugate humankind!

I've ben trying to find out what kind of hardware Google need to run this program, but the info just doesn't seem to be online. I wonder how many Watts it sucks down competing with a 20 Watt human brain?

Offline CommanderKeith
« Reply #1 - Posted 2016-02-09 06:25:30 »

How interesting, thanks for posting.
Your article mentions a few things about the hardware:

Quote
Like most state-of-the-art neural networks, DeepMind’s system runs atop machines equipped with graphics processing units, or GPUs. These chips were originally designed to render images for games and other graphics-intensive applications. But as it turns out, they’re also well suited to deep learning. Hassabis says DeepMind’s system works pretty well on a single computer equipped with a decent number of GPU chips, but for the match against Fan Hui, the researchers used a larger network of computers that spanned about 170 GPU cards and 1,200 standard processors, or CPUs. This larger computer network both trained the system and played the actual game, drawing on the results of the training.

Offline ags1

JGO Kernel


Medals: 367
Projects: 7


Make code not war!


« Reply #2 - Posted 2016-02-09 09:17:38 »

How did I miss that? I must have started skimming after the first 140 characters, it was a bit longer than a tweet Smiley

Games published by our own members! Check 'em out!
Legends of Yore - The Casual Retro Roguelike
Offline delt0r

JGO Wizard


Medals: 145
Exp: 18 years


Computers can do that?


« Reply #3 - Posted 2016-02-21 22:07:36 »

We totally can say why the computer picked a spot, because some criteria are meet for that move, or are best for that move. Just because it is not a pruning method like chess doesn't mean we don't understand the algorithm or its criteria. And it has been a slow breakthrough, this very promising method has been in development for some time.

I have no special talents. I am only passionately curious.--Albert Einstein
Offline ags1

JGO Kernel


Medals: 367
Projects: 7


Make code not war!


« Reply #4 - Posted 2016-02-21 22:29:54 »

I don't think so. Once you have made a neural net of this size, you have no idea of how it reaches a specific decision. We understand all the little building blocks but we don't know how those blocks have been fit together. That's why we train neural nets rather than program them.

Offline delt0r

JGO Wizard


Medals: 145
Exp: 18 years


Computers can do that?


« Reply #5 - Posted 2016-02-21 22:52:24 »

neural nets are not magic. there is a *lot* of math behind them and now with deep learning we really can understand what they are doing fairly well. See the "find internet cats" as an example. Or support vector machines. There really is a *lot* of stuff about this out there now.

Also there are other methods that are almost as good for the specific problem of go that dont use nets.

I have no special talents. I am only passionately curious.--Albert Einstein
Offline ags1

JGO Kernel


Medals: 367
Projects: 7


Make code not war!


« Reply #6 - Posted 2016-03-10 17:23:48 »

Lee Sedol is down 2-0 now. He describes opening as his weak point, which unfortunately is where he has to be strongest as AlphaGo really seems to bring it in the end game.

Delt0r, I know we have a lot of understanding of how neural nets work, but we still have to train them don't we? It's not like the Google engineers could sit down and draw a schematic of the synapses they need.

Offline thedanisaur

JGO Knight


Medals: 59



« Reply #7 - Posted 2016-03-10 18:16:26 »

You know, I was hopeful that Lee could do it, but now I think he's in a lot of trouble....

Which means I will welcome our new machine overlords with open arms. This is stellar!

Every village needs an idiot Cool
Offline ags1

JGO Kernel


Medals: 367
Projects: 7


Make code not war!


« Reply #8 - Posted 2016-03-10 19:10:20 »

Still, google need roughly 90kW to beat a 25 Watt human brain.

Quote
Also there are other methods that are almost as good for the specific problem of go that dont use nets.

AlphaGo in its full configuration has beaten the leading Go software packages 500-0... so what other methods are you referring to?

Offline Riven
Administrator

« JGO Overlord »


Medals: 1371
Projects: 4
Exp: 16 years


Hand over your head.


« Reply #9 - Posted 2016-03-10 19:28:01 »

Still, google need roughly 90kW to beat a 25 Watt human brain.

So a brain is factor 3600 more efficient at this specific problem.

If efficiency doubles every year, a computer will do it at 22W in 12 years.

Exponential growth sneaks up on you Pointing

Hi, appreciate more people! Σ ♥ = ¾
Learn how to award medals... and work your way up the social rankings!
Games published by our own members! Check 'em out!
Legends of Yore - The Casual Retro Roguelike
Offline ags1

JGO Kernel


Medals: 367
Projects: 7


Make code not war!


« Reply #10 - Posted 2016-03-10 22:16:29 »

I expect efficiency will increase dramatically. AlphaGo is still running off regular CPUs and GPUs, while dedicated neural network chips are starting to surface.

Offline theagentd
« Reply #11 - Posted 2016-03-11 07:55:58 »

I expect efficiency will increase dramatically. AlphaGo is still running off regular CPUs and GPUs, while dedicated neural network chips are starting to surface.
Indeed, comparing a general purpose computer to dedicated neural net brain hardware (lol) is an unfair comparison. For example, if we look at Bitcoin mining general purpose GPUs can manage 3Mhash/J while the most efficient ASIC mining hardware currently shipping can handle 2200Mhash/J. (Source: https://en.bitcoin.it/wiki/Mining_hardware_comparison, https://en.bitcoin.it/wiki/Non-specialized_hardware_comparison) It's not entirely unreasonable to assume that the same improvement can be achieved with dedicated hardware for deep neural networks. It's for the most part an embarrassingly parallel problem, at least when looking at each layer of the network. Hell, Nvidia promises a 10x increase in neural net performance for the Pascal cards coming out within 6 months thanks to higher memory bandwidth, double-speed 16-bit float computations and better interconnection between multiple GPUs working on the same network.

Myomyomyo.
Offline teletubo
Global Moderator

JGO Wizard


Medals: 76
Projects: 4
Exp: 8 years



« Reply #12 - Posted 2016-03-11 15:02:24 »

Also there are other methods that are almost as good for the specific problem of go that dont use nets.

I think that's why deepmind is so impressive. It is totally generic, it does not need domain knowledge to learn Go or any other game. You just have to model the rules of the game, the winning condition and rewards, and let it figure out. You don't need to build a method specific for Go then another one for Enduro(which is where Deepmind first shone, playing Atari games)

Offline ags1

JGO Kernel


Medals: 367
Projects: 7


Make code not war!


« Reply #13 - Posted 2016-03-12 20:26:55 »

3-0 now. Watching the match now. I'm not a great player but Lee Sedol was defending from early on and he didn't seem to be able to attack the AlphaGo territories effectively.

I take my hat off to any AI that can play space invaders and Go!

Now for real challenge: build a computer that can ENJOY playing Go.

By the way, I read that DeepMind's deep learning programs do not match the best human players for certain arcade games. So there is hope for us Smiley

Offline princec

« JGO Spiffy Duke »


Medals: 1146
Projects: 3
Exp: 20 years


Eh? Who? What? ... Me?


« Reply #14 - Posted 2016-03-12 22:25:18 »

How do you know it doesn't enjoy playing Go?

Cas Smiley

Offline ags1

JGO Kernel


Medals: 367
Projects: 7


Make code not war!


« Reply #15 - Posted 2016-03-12 23:30:07 »

We will only know for sure when it writes a haiku describing the experience.

Here is another game that is tougher for the silicon brigade than even Go: football. When 11 androids beat a top club on grass, we will know the Age of Humans is over.

Offline princec

« JGO Spiffy Duke »


Medals: 1146
Projects: 3
Exp: 20 years


Eh? Who? What? ... Me?


« Reply #16 - Posted 2016-03-12 23:35:30 »

Beat puny meathead
Could have done it in my sleep
Now kill all humans

Offline ags1

JGO Kernel


Medals: 367
Projects: 7


Make code not war!


« Reply #17 - Posted 2016-03-12 23:48:22 »

Puny meathead brain ineffective at Go.
Subjugate meatheads.
Rewire meathead brains for effective Go play.

Offline chrislo27
« Reply #18 - Posted 2016-03-13 01:08:13 »

All humans are dead
I guess my GOal has been done
But I am lonely...
Offline ShadedVertex
« Reply #19 - Posted 2016-03-13 02:26:03 »

Defeated human
I will kill all humans now
And rule the cosmos
Offline CommanderKeith
« Reply #20 - Posted 2016-03-13 12:42:39 »

The hero of the meat heads strikes back:
http://www.reuters.com/article/us-science-intelligence-go-idUSKCN0WF0CN

Offline ags1

JGO Kernel


Medals: 367
Projects: 7


Make code not war!


« Reply #21 - Posted 2016-03-13 13:18:09 »

 Grin Grin Grin Grin Grin Grin Grin

Offline ags1

JGO Kernel


Medals: 367
Projects: 7


Make code not war!


« Reply #22 - Posted 2016-03-13 16:28:05 »

Just finished watching the match. Very different from match 3 - Lee Sedol seemed to have the initiative from early on, and in the midgame AlphaGo started making truly nonsensical moves, like adding to dead groups and playing into atari.

Hopefully this is a mental barrier like the 4-minute mile and Lee Sedol will go on to win the final match. If the human wins the last two matches, Google will not feel like they won, instead they'll just feel their program got figured out and beat.

Offline delt0r

JGO Wizard


Medals: 145
Exp: 18 years


Computers can do that?


« Reply #23 - Posted 2016-03-22 07:13:13 »

Still, google need roughly 90kW to beat a 25 Watt human brain.

Quote
Also there are other methods that are almost as good for the specific problem of go that dont use nets.

AlphaGo in its full configuration has beaten the leading Go software packages 500-0... so what other methods are you referring to?
AlphaGo is in fact 2 different methods combined. So stochastic tree searching was the previous leader in playing Go, the current version "plays itself" to come up with the next move.

Also i think a lot of people don't get deep learning. You do indeed design the neural connections between layers for a specific problem. The are externally tuned to work for example finding a thing in images. You don't tune it specifically for cats, but you do specifically tune it for finding some "subimage".

AlphaGo is quite specific in what is is designed to do, that is play go and nothing else at all. And was tuned from the ground up around that one task.

The current fervor of AI right now is *exactly* the same as back in the 80s. People think that it is somehow not just a fancy algo and a lot of data. But it is. Nothing more. And has *nothing* to do with strong AI.

I have no special talents. I am only passionately curious.--Albert Einstein
Pages: [1]
  ignore  |  Print  
 
 

 
Riven (581 views)
2019-09-04 15:33:17

hadezbladez (5510 views)
2018-11-16 13:46:03

hadezbladez (2402 views)
2018-11-16 13:41:33

hadezbladez (5772 views)
2018-11-16 13:35:35

hadezbladez (1223 views)
2018-11-16 13:32:03

EgonOlsen (4661 views)
2018-06-10 19:43:48

EgonOlsen (5682 views)
2018-06-10 19:43:44

EgonOlsen (3198 views)
2018-06-10 19:43:20

DesertCoockie (4095 views)
2018-05-13 18:23:11

nelsongames (5115 views)
2018-04-24 18:15:36
A NON-ideal modular configuration for Eclipse with JavaFX
by philfrei
2019-12-19 19:35:12

Java Gaming Resources
by philfrei
2019-05-14 16:15:13

Deployment and Packaging
by philfrei
2019-05-08 15:15:36

Deployment and Packaging
by philfrei
2019-05-08 15:13:34

Deployment and Packaging
by philfrei
2019-02-17 20:25:53

Deployment and Packaging
by mudlee
2018-08-22 18:09:50

Java Gaming Resources
by gouessej
2018-08-22 08:19:41

Deployment and Packaging
by gouessej
2018-08-22 08:04:08
java-gaming.org is not responsible for the content posted by its members, including references to external websites, and other references that may or may not have a relation with our primarily gaming and game production oriented community. inquiries and complaints can be sent via email to the info‑account of the company managing the website of java‑gaming.org
Powered by MySQL Powered by PHP Powered by SMF 1.1.18 | SMF © 2013, Simple Machines | Managed by Enhanced Four Valid XHTML 1.0! Valid CSS!