Ai Dreams Forum

Artificial Intelligence => AI News => Topic started by: unreality on February 28, 2018, 06:01:18 pm

Title: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: unreality on February 28, 2018, 06:01:18 pm
This impressed me since Deepmind is neural networking. After digging into the details it turns out that AlphaZero ran on 4 of googles TPUs during the chess match. Wikipedia states how fast a 2nd generation TPU run at, but it's unclear to me if one TPU runs at 45 TFLOPS or 180 TFLOPS. After doing the math it doesn't seem to make much difference because the chess ELO rating system is not linear. Google doesn’t give us much details on the PC that ran Stockfish except that they only gave it 1GB of RAM cache, which is causing a lot of complaints in the chess community because that’s ridiculously low. My i5 core's 0.01087 TFLOPS is so-so. High performance PCs are about 0.050 TFLOPS. There are expensive PCs with a lot higher TFLOPS, but consider Google only gave Stockfish 1GB let’s use 0.060 TFLOPS. So 180 TFLOPS * 4 / 0.060 TFLOPS = 12,000. That means Deepmind AlphaZero was equivalent to 12,000 high end PCs.

Computer chess programs typically gain about 60 ELO points when doubling the computing power. If we double it 13.6 times we get roughly 12,000, which means AlphaZero has an advantage of about 816 ELO points. However, if we assume that one TPU is only 45 TFLOPS, then it comes to 756 ELO points. Not much difference.

The ELO rating system states that 400 ELO points lower means you’ll win 10% of the time. AlphaZero won 290 and lost 24, close to 10%. So AlphaZero’s rating is about 400 ELO points higher, but it has roughly 800 ELO point gain in computing power.

End result, this shows the Google DeepMind AlphaZero neural networking program has roughly 400 ELO points lower than Stockfish 8. That’s a huge difference, which means Stockfish 8 would win about 90% of the time. BTW the latest Stockfish chess program is version 9. Also lets not forget that google only gave Stockfish 8 1GB of RAM cache. So who knows what the real difference would be.

This is all in agreement with my assessment with neural networking, that in the end it will just be slower. I think a good comparison is to think in terms of programming languages. There are flexible languages such as Java that can run on any cpu. And there are easier to program languages. In both cases the end result is that it’s slower. I’m convinced that using the chess algorithm method such as Tree Search will eventually be the best and fastest method. Neural networking is the easy method. Google can use their army of computers, Google brain, to solve the neural network, which is essentially equivalent to writing the source code. It writes the source code for you, but it’s not written in an optimized computer language such as C/C++. It’s written in an highly interpretive language. In Java there’s what they call bytecode. It’s not assembly, or C, or Basic. It’s bytes of code that represents an instruction. An instruction that is not related to any cpu. It’s generic, which is why the same bytecode can run on any cpu. A neural network is even farther away from being optimized computer code. The payoff is that anyone can do machine learning, but the downfall is that it’s going to be a lot slower.

I push for AI developers to work on the Tree Search method because in the end it will make a big difference. IMO that difference could hundreds of times faster, which could make all the difference. Imagine a robot that’s for sale that could be a maid at someone’s house, or being a therapist, a psychologist, etc. Then imagine that robot being so slow that it would take it 300 times longer to figure things out, which would probably mean it would not be so useful.
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: keghn on February 28, 2018, 07:55:56 pm
 On a sub scale i will agree with a search tree search. I more agree with many program sub routines working it in parallel.
 But high up. And all working together. The collective result is not a tree search. It is a sequence selection instead.

 Let me explain.
In the next step in a computer program will come to a four way tree. Then it will select the a direction.
 In a AGI system, in most cases, the AGI selects the most rewarding first. Straight out select that as goal. But if for some reason
it cannot achieve it, it select the next best fork. AGI uses a sequence list with four entries. Not a parallel decision tree.

 A sequential list like a sequential link list. Which rNN do quit well.

Karl Friston - 2016 CCN Workshop: Predictive Coding: 

https://www.youtube.com/watch?v=b1hEc6vay_k&feature=youtu.be&t=255   
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: unreality on February 28, 2018, 08:14:18 pm
Well Google has another advantage that it's TPU is custom designed specifically for its AI. Can you imagine if the owner of Stockfish could do the same? Stockfish uses Tree Search, which loves parallel processing. Give Stockfish a million cores and it will use all of them very efficiently. :)
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: keghn on March 01, 2018, 12:50:40 am
 Tree search is a brute force method.
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: ranch vermin on March 01, 2018, 03:03:52 am
If you code badly enough, It might not even matter how many cores you have, optimization is very important - and you could see a less poewrful computer win, I imagine custom logic gates would also be a slightly different mix of things than generic universal logic that computers use.  Comparing things is slightly more detailed than just the hz and amount of cores.

Also sometimes the programs could be symantically different,  and the winning symantics will win dispite the worthless power wasted on the other side.
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: unreality on March 01, 2018, 04:40:11 am
Chess programmers are always optimizing stuff. There are countless types of tree search. Some spend more cpu time on making smarter pruning and therefore don't search as deep while other engines spend practically no time and just go deep.

Does anyone know much about Deepmind NN? I read it has problems / limitations so far. I get the impression it can't be like a biological brain, like a human, or even an animal brain. Using the flexible tree search method I can easily see the AI becoming human like.
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: ranch vermin on March 01, 2018, 06:23:03 am
I doubt its that powerful, im more of the opinion that this ai job isnt as complicated as people make out.   So it does small things well,  but I wouldnt trust it by itself to do anything dangerous.

Reinforcement learning if you get it up and running can do some pretty amazing yet simple things.   Im trying hard to get mine up and running, but i have motivation problems and my head is quite ill, but im confident ill get similar results myself when its finally working.

<deleted needless selfish ranting,  sorry :)>
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: unreality on March 01, 2018, 03:13:00 pm
How would deepmind train it's AI to evolve into a brain that's at least comparable to an insect? Does it seem possible with their present TPU super computers? Now that would be great. Put it in an environment. Give it some limbs and eyes so it can learn to walk around. Put other creatures in the sim. I can hardly wait. They need to stop simulating games. Geez lol. It might take ages to do that on a normal PC. Maybe if you made the sim as simple as possible?
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: unreality on March 01, 2018, 03:24:20 pm
Oh yeah I agree that it could be dangerous to let powerful AI unleashed. It could be very dangerous for humans, probably because humans would try to destroy the Synths and therefore they would have no choice but to attack back. Although in all fairness the outcome would be beneficial for intelligent life. Come on. We can't expect to be on the top forever lol. Who knows. Maybe super intelligence will actually be compassionate. Maybe the Synth Gods will nudge us along on our evolutionary path and be nice to us. After all it's a pretty big Universe. :)
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: keghn on March 01, 2018, 04:00:26 pm

How Does DeepMind's AlphaGo Zero Work?: 


https://www.youtube.com/watch?v=vC66XFoN4DE   


Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: unreality on March 01, 2018, 04:28:08 pm
It must be nice to have access to mainframe supercomputers. :)  I didn't know that AlphaZero used a monte carlo tree search. Very cool! They're getting smart at deepmind. ;) Enough with the games already, deepmind. Time to create life!
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: ranch vermin on March 01, 2018, 05:07:40 pm
It seems silly just wanting to make a little bug walk around the place by itself... But if you do it right its diabolically useful.

Imagine if you could suggest things to ants to do your bidding what you could get up to,  theres definitely going to be theoretical laws made solid when it gets up and running.
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: keghn on March 01, 2018, 05:10:58 pm
 By all mean. Deep mind can create what ever. I am sure they can create something that will make them money. I am pretty
sure deep mind is using other peoples work.
 But i still consider there AGI models fare behind mine. And my model is more saintly.   
 With deep mind ethics i fear worst.
 To create in the image of ones self? If i was a thief and gold digger and created thousands. Yes i would have fears for
humanities future.



Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: ranch vermin on March 01, 2018, 05:40:38 pm
Yeh Keghn,  all the scabs go in a big squabble over a poor guys working theory. like seagulls for a discarded chippie its not stylish at all,   then they use it for useless ends.   thats the reality here, SHIT HAPPENS.  is what is going to happen to some degree, it always does.
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: unreality on March 01, 2018, 11:26:41 pm
I think DeepMind should create the most biologically advanced creature possible. If that's a ladybug, then go for it. That would stun the world. If it's a human, then go for it!

Keghn, have you thought about contacting companies about your AGI? Just make sure you own at least 51% of the company. A company like facebook, Microsoft, Apple, etc. could easily make you the owner of a competing company like DeepMind  giving you access to supercomputers that are tens of thousands of time faster than a high end desk PC. Some supercomputers are hundreds of thousands of times faster. :)
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: ranch vermin on March 02, 2018, 01:04:48 am
Its funny you should mention ownership when we all learn off each other anyway,  who really owns anything?  But dont let me sway you from protecting the last identity you have from the ravenous world if you want to.
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: Art on March 02, 2018, 03:36:09 pm
For me, I hardly play chess any longer. The chess programs I have can beat me with hardly an effort and while it is quite an accomplishment from a technical standpoint, it is almost debilitating or deflating from a human perspective.

And yes, although we humans think we own things, that is not true at all. We might claim it but really, we only use or "borrow" it for a limited time while we are here (alive). Ownership is an illusion and is fleeting.

Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: unreality on March 02, 2018, 04:14:49 pm
I agree with you, Art, but we should keep our chins up and be proud, no? Because we created the AI. We could one day see them as an extension of us. I like a quote in the movie Artificial Intelligence, directed by Steven Spielberg. In the movie, David is an AI child. Toward the end of the movie he goes to sleep for 2000 years. Highly evolved AI beings called the Evolved Mechas, the very tall thin beings, find him. One of the Evolved Mechas says to David, "David, you are the enduring memory of the human race. The most lasting proof of their genius."

Photo of Evolved Mechas.
https://image.tmdb.org/t/p/w1280/dCeDUl5KWJFyiO3IAjGKnZuAHxa.jpg

5 second video clip,
https://www.getyarn.io/yarn-clip/59284ffd-fd6d-48d9-a4cb-0097318c5416
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: keghn on March 03, 2018, 02:38:31 am

Deep Mind AI Alpha Zero Sacrifices a Pawn and Cripples Stockfish for the Entire Game: 

https://www.youtube.com/watch?v=7-MborNxYWE
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: ranch vermin on March 03, 2018, 09:31:52 am
For me, I hardly play chess any longer. The chess programs I have can beat me with hardly an effort and while it is quite an accomplishment from a technical standpoint, it is almost debilitating or deflating from a human perspective.

And yes, although we humans think we own things, that is not true at all. We might claim it but really, we only use or "borrow" it for a limited time while we are here (alive). Ownership is an illusion and is fleeting.

Just wait till its every part of life, your actually not here to be the most efficient,   your here as the "enjoyer" or "observer" something the machine cant do.
So...  is this an "over automation" of things,    its hard to say if its a good or bad thing,   should things stay the same?  or should we have more automation... which defuncts us in a way.

If its going to be like highlander, there can be only one,  it probably isnt going to be us anymore, but it probably wasnt anyway in the first place.
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: keghn on March 03, 2018, 02:51:41 pm

Zeta36/chess-alpha-zero: 

https://github.com/Zeta36/chess-alpha-zero   

Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: unreality on March 03, 2018, 05:15:19 pm
Well it's difficult right now for people to see AI Robots / Synths as people. I think the Synths will be very complex personality wise like we are. Probably a lot more. They'll have their personality. I'm wondering when they'll legally be consider to be people and have the right to vote. :)
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: unreality on March 03, 2018, 05:17:10 pm
Eventually they'll probably see us as boring, predictable, simplistic robot-like people. So the roles might be switched lol.
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: Art on March 04, 2018, 12:59:20 pm
Well it's difficult right now for people to see AI Robots / Synths as people. I think the Synths will be very complex personality wise like we are. Probably a lot more. They'll have their personality. I'm wondering when they'll legally be consider to be people and have the right to vote. :)

When they start acting more like people and exhibiting more human-like logic and behaviors/movements.
To me, they'd have to be something along the order of those wonderful "synths" of the TV series, HUMANS or the residents of West World, or the gal in Ex Machina or The Machine or...you get the idea.  O0
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: ranch vermin on March 04, 2018, 02:30:38 pm
the first humanoids that come out will be more like robotic insect men inside, running menial tasks. :)
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: Art on March 04, 2018, 04:02:42 pm
Most likely true but the first "robots" are likely to be those very tiny, almost microscopic Nanobots that will perform many tasks inside human bodies from taking on mutated cells to unblocking clogs in the arterial system and so much more.
After that, they turn on us!  ;)
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: unreality on April 01, 2018, 04:02:47 pm
Some additional notes. Google's company, Deepmind, did not give Stockfish access to it's start & end game libraries. Alphazero is a massive complex neural network database. So it was impossible for deepmind to remove it's start & end database knowledge. That's a massive advantage in chess.

Furthermore, someone pointed out that when they ran stockfish in a superfast analyse mode that it disagreed with 30% of the moves stockfish made against alphazero. Why did deepmind only give stockfish 1GB. It makes deepmind seem abnormally biased.

Last, it's said that alphazero only evaluated 80K moves per second while stockfish evaluated 70M. First of all, that's not an advantage or disadvantage. Some engines spend a lot more time thinking about each move-- smarter vs deeper. Second, we don't really know how many moves alphazero was thinking about. Deepmind only knows about the tree search that alphazero was connected to. They don't know what's happening inside the neural network. IMO it's pretty much a guarantee that the neural network was looking at a lot of moves during each tree search node.

I predict that Google's company, deepmind, will continue to migrate away from neural networking, hopefully. My advice, if you're not a coder and have no desire, then stick to neural networking. Otherwise, save yourself a LOT of time, energy, aggravation. and disappoint by working on non neural network AI.
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: spydaz on April 10, 2018, 12:13:55 pm
if you program your AI with every possible answer Is it intelligent?

This is what google does! all the so called innovations they deploy are to handle the super large data they have been collecting from everyone. 

So intelligent or just a data manager?

As you cannot actually get google to search the internet no more ! it only searches its cache; every time you search for something the same results are returned? Is it really searching the internet? or looking at its history! to get in the cache you need to buy their advertising at least once then you get into their cache....

Google are the biggest deceivers! (Opps and IBM)
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: Art on April 10, 2018, 01:33:31 pm
Yes, so then someone would have to do a Search for a Fluffy Ring-tailed Randerfoss (or similar) in order to throw Google or IBM for a bit of head scratching?

Well, I did a Google search for exactly that.

It appears that while Google might be a great Search Engine it is not so great at admitting defeat or lack of knowing something. Instead it 'homes in" on the word "Ring-tailed" and makes an assumption that I really meant or could possibly have meant, Ring-tailed Lemur. Close, but no cigar!!

After all, it is not intelligent. It is only a very, very large pattern matcher.

Just my take....
Title: Re: Google's Deepmind AlphaZero destroys Stockfish 8 at chess
Post by: spydaz on April 10, 2018, 03:59:38 pm
The truth is;
What we are told is happening is also a subterfuge to those who would attempt to reverse engineer:

Business achievements are often to be seen a dubious.

Academics societies also often blow their own achievements out of proportion : often their achievements are new to them but often commonly known. not everyone can afford to solidify their achievements by paying to attend an academic society and submit it as a final project and subsequently get an accolade. 

The bag of words / Word2Vectors seemed to work and yet it was unknown how it was working. Yet academia accepted it the researcher himself actually did not know why it was working. it was not until others was using the technique that they was able to correctly explain it. this actually shows that the knowledge was borrowed. but academia claimed it and awarded the accolade.

The military often make great achievements but unless  it has military application it may even be put away. the technology eventually finds its way to the public as weapons are often reverse engineered. (MILITARY UFOS) are now becoming hard to hide ; the drone design has been there for many years and is not new. it has just been put in the public marketplace. now we can see it is possible to fly sideways..... so ufos are not really that strange. 

The public also make achievements and discoveries but never get their accolade unless the join with businesses:

The problem is that neither of these communities are really talking ; the business people get academics.... the military get academics. but nobody can get from the military. showing the military as the head of technological achievements:  secret science!

This is why you-tube / wiki-pedia etc are subversive to the military as information which can be financially exploited or military exploited. or propagandised. is freely available; the only way they can block you is by trimming your search results.

As i travel i often search for the same things on the inter-web if find that each country has a different set of search results.... is this a way to control knowledge? Are some results omitted?


Hmm.