World #1 Go Player Ke Jie accepts Google Alpha Go Match..

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
duncan
Posts: 11173
Joined: Mon Jul 07, 2008 8:50 pm

Re: World #1 Go Player Ke Jie accepts Google Alpha Go Match.

Post by duncan » Tue May 23, 2017 11:22 am

Laskos wrote:
Michel wrote:Only by 0.5 point if I understood correctly. On the other hand AlphaGo is programmed to go for the surest win and not the win that delivers the most points (which may involve taking higher risks). So the fact that AlphaGo "only" won by 0.5 points may be misleading.
0.5 without komi of 7.5? That was my impression, but I don't know how to count very well. From what I understood, Ke Jie tried to gain points, so that Alpha would need to attack, but mishandled the opening and in fact after the opening Alpha had more (potential) points and Ke had to attack. There was no opportunity to attack Alpha in the whole game, and Alpha, gaining advantage from the opening, just played quietly for the win. Alpha still seems to go to the highest probability, not highest advantage, so potentially it may had 10-15 stones advantage in late midgame, but went for the safer 6-8 points. I will maybe post Crazy Stone Deep Learning (7 amateusr dan) 2 hour analysis, that shows that from this engine (Zenith 6 too) point of view, White (Alpha) had substantial advantage right from the opening after move 20, a thing which seems to be overlooked or muted by 9p pros commenting. It seems that even relatively weak engines compared to those pros (3-4 stones difference) are better at scoring in the initial stages of the game.

Ke here seems to have had no any chances, a pretty overwhelming win by Alphago.
thanks for the analysis. keep it going!


with chess most (although not all) experts seem to be saying even with a 32 piece database, a computer with knight handicap would not be able to defeat a grandmaster.

do you have an opinion what handicap would be needed for a 9 dan go player to defeat a computer with access to a solved database of go.?

User avatar
Laskos
Posts: 10181
Joined: Wed Jul 26, 2006 8:21 pm
Full name: Kai Laskos

Re: World #1 Go Player Ke Jie accepts Google Alpha Go Match.

Post by Laskos » Tue May 23, 2017 3:29 pm

duncan wrote:
Laskos wrote:
Michel wrote:Only by 0.5 point if I understood correctly. On the other hand AlphaGo is programmed to go for the surest win and not the win that delivers the most points (which may involve taking higher risks). So the fact that AlphaGo "only" won by 0.5 points may be misleading.
0.5 without komi of 7.5? That was my impression, but I don't know how to count very well. From what I understood, Ke Jie tried to gain points, so that Alpha would need to attack, but mishandled the opening and in fact after the opening Alpha had more (potential) points and Ke had to attack. There was no opportunity to attack Alpha in the whole game, and Alpha, gaining advantage from the opening, just played quietly for the win. Alpha still seems to go to the highest probability, not highest advantage, so potentially it may had 10-15 stones advantage in late midgame, but went for the safer 6-8 points. I will maybe post Crazy Stone Deep Learning (7 amateusr dan) 2 hour analysis, that shows that from this engine (Zenith 6 too) point of view, White (Alpha) had substantial advantage right from the opening after move 20, a thing which seems to be overlooked or muted by 9p pros commenting. It seems that even relatively weak engines compared to those pros (3-4 stones difference) are better at scoring in the initial stages of the game.

Ke here seems to have had no any chances, a pretty overwhelming win by Alphago.
thanks for the analysis. keep it going!


with chess most (although not all) experts seem to be saying even with a 32 piece database, a computer with knight handicap would not be able to defeat a grandmaster.

do you have an opinion what handicap would be needed for a 9 dan go player to defeat a computer with access to a solved database of go.?
Go is much deeper than Chess, also visible in the ELO span from patzer to super-expert, about 2000 ELO points in Chess, about 5000 ELO points in Go. Go is also probably less studied than Chess. One can see that this AlphaGo is outplaying best humans in the opening, where one would expect humans excel. Contrary to Chess (well, the approaches are also different, A/B compared to MC and DCNN). I would guess the handicap should be huge, in dozen handicap stones, but I am a complete patzer in Go. Go could have tens of thousands ELO points to improve, each handicap stone now at top level is about 200-300 ELO points, but its price increases with level of play, so it's hard to gauge the number of handicap stones to the perfect play. Also, it depends on komi (7.5 already seems a bit much to engines) and rules used (Chinese, Japanese).

User avatar
Laskos
Posts: 10181
Joined: Wed Jul 26, 2006 8:21 pm
Full name: Kai Laskos

Re: World #1 Go Player Ke Jie accepts Google Alpha Go Match.

Post by Laskos » Tue May 23, 2017 3:52 pm

lkaufman wrote:
Dirt wrote:The match is now scheduled to take place from May 23 to May 27 of 2017. Only three games are in the main match but there will also be matches between professionals alternating moves with computers as well as other activities.
Yes, it starts tonite U.S. time. Michael Redmond 9 Dan pro estimates a 10% chance of the pro winning even 1 game. He thought that a 2 stone handicap would be fair, but "politically incorrect".
By a curious coincidence (?) just last week the top Shogi engines defeated the World Champion ("Meijin" to be exact) Sato 2 to 0 in comparably serious games, and they weren't even close; my judgment as an amateur 5 Dan is that the engine could easily give a Lance handicap to any human and expect to win So assuming that the computer wins this Go match, Go will have lasted one week longer than Shogi before having fallen to the computers!
This thing which destroyed Ke Jie runs on a machine 10 times smaller than that against Lee Sedol, a large single machine with like 4 GPU. I would dare to say that AlphaGo on home PC is stronger than top Go pros. A remarkable achievement. They also have new approaches built in, ready to be published in "Nature" after the match. I guess in 3-4 years we will have Go on Smartphone beating top Go players. Outstanding progress, this seemed very far away only 2 years ago.

Dirt
Posts: 2851
Joined: Wed Mar 08, 2006 9:01 pm
Location: Irvine, CA, USA

Re: World #1 Go Player Ke Jie accepts Google Alpha Go Match.

Post by Dirt » Tue May 23, 2017 3:55 pm

duncan wrote:with chess most (although not all) experts seem to be saying even with a 32 piece database, a computer with knight handicap would not be able to defeat a grandmaster.

do you have an opinion what handicap would be needed for a 9 dan go player to defeat a computer with access to a solved database of go.?
It's an ill posed question.

In chess, what do you do when the tablebase says you are losing? You might try to maximize the moves to mate but that probably isn't the best strategy to win against an imperfect player. In fact, the strategy will depend on that player. Go is probably similar.
Deasil is the right way to go.

Milos
Posts: 3534
Joined: Wed Nov 25, 2009 12:47 am

Re: World #1 Go Player Ke Jie accepts Google Alpha Go Match.

Post by Milos » Tue May 23, 2017 4:44 pm

Laskos wrote:This thing which destroyed Ke Jie runs on a machine 10 times smaller than that against Lee Sedol, a large single machine with like 4 GPU. I would dare to say that AlphaGo on home PC is stronger than top Go pros. A remarkable achievement. They also have new approaches built in, ready to be published in "Nature" after the match. I guess in 3-4 years we will have Go on Smartphone beating top Go players. Outstanding progress, this seemed very far away only 2 years ago.
I wouldn't be so sure about that. The difference between Titan Xp and Adreno 540 is far greater than difference between i7-6950X and Snapdragon 835.

User avatar
Laskos
Posts: 10181
Joined: Wed Jul 26, 2006 8:21 pm
Full name: Kai Laskos

Re: World #1 Go Player Ke Jie accepts Google Alpha Go Match.

Post by Laskos » Tue May 23, 2017 4:57 pm

Milos wrote:
Laskos wrote:This thing which destroyed Ke Jie runs on a machine 10 times smaller than that against Lee Sedol, a large single machine with like 4 GPU. I would dare to say that AlphaGo on home PC is stronger than top Go pros. A remarkable achievement. They also have new approaches built in, ready to be published in "Nature" after the match. I guess in 3-4 years we will have Go on Smartphone beating top Go players. Outstanding progress, this seemed very far away only 2 years ago.
I wouldn't be so sure about that. The difference between Titan Xp and Adreno 540 is far greater than difference between i7-6950X and Snapdragon 835.
In many benchamrks the latest Samsungs are only 5 time slower than a fully equipped Intel 7700K. The progress in Go was 2000+ ELO points in the last 1.5 years. Add another 1000 in 3-4 years and some moderate progress of Samsung, and you are there. Let's see, but I cannot wait to buy the new Zen for PC, or even better, AlphaGo if they make it commercial. I already bought a strong Nvidia card. Stupid hobbies, as a player I am a patzer in everything :).

Milos
Posts: 3534
Joined: Wed Nov 25, 2009 12:47 am

Re: World #1 Go Player Ke Jie accepts Google Alpha Go Match.

Post by Milos » Tue May 23, 2017 5:07 pm

Laskos wrote:
Milos wrote:
Laskos wrote:This thing which destroyed Ke Jie runs on a machine 10 times smaller than that against Lee Sedol, a large single machine with like 4 GPU. I would dare to say that AlphaGo on home PC is stronger than top Go pros. A remarkable achievement. They also have new approaches built in, ready to be published in "Nature" after the match. I guess in 3-4 years we will have Go on Smartphone beating top Go players. Outstanding progress, this seemed very far away only 2 years ago.
I wouldn't be so sure about that. The difference between Titan Xp and Adreno 540 is far greater than difference between i7-6950X and Snapdragon 835.
In many benchamrks the latest Samsungs are only 5 time slower than a fully equipped Intel 7700K. The progress in Go was 2000+ ELO points in the last 1.5 years. Add another 1000 in 3-4 years and some moderate progress of Samsung, and you are there. Let's see, but I cannot wait to buy the new Zen for PC, or even better, AlphaGo if they make it commercial. I already bought a strong Nvidia card. Stupid hobbies, as a player I am a patzer in everything :).
As I said i7-6950X vs Snapdragom 835 is like 5x for chess (SF benchmark).
However TitanXp is 30x compared to Adreno 540 and that's in games, and in actual ML difference is easily 50x. And AlphaGo hardware atm has 4 TitanXp's. That's close to 200x and AlphaGo strength lies almost exclusively in GPU power.

David Xu
Posts: 47
Joined: Mon Oct 31, 2016 8:45 pm

Re: World #1 Go Player Ke Jie accepts Google Alpha Go Match.

Post by David Xu » Tue May 23, 2017 5:23 pm

Milos wrote:
Laskos wrote:
Milos wrote:
Laskos wrote:This thing which destroyed Ke Jie runs on a machine 10 times smaller than that against Lee Sedol, a large single machine with like 4 GPU. I would dare to say that AlphaGo on home PC is stronger than top Go pros. A remarkable achievement. They also have new approaches built in, ready to be published in "Nature" after the match. I guess in 3-4 years we will have Go on Smartphone beating top Go players. Outstanding progress, this seemed very far away only 2 years ago.
I wouldn't be so sure about that. The difference between Titan Xp and Adreno 540 is far greater than difference between i7-6950X and Snapdragon 835.
In many benchamrks the latest Samsungs are only 5 time slower than a fully equipped Intel 7700K. The progress in Go was 2000+ ELO points in the last 1.5 years. Add another 1000 in 3-4 years and some moderate progress of Samsung, and you are there. Let's see, but I cannot wait to buy the new Zen for PC, or even better, AlphaGo if they make it commercial. I already bought a strong Nvidia card. Stupid hobbies, as a player I am a patzer in everything :).
As I said i7-6950X vs Snapdragom 835 is like 5x for chess (SF benchmark).
However TitanXp is 30x compared to Adreno 540 and that's in games, and in actual ML difference is easily 50x. And AlphaGo hardware atm has 4 TitanXp's. That's close to 200x and AlphaGo strength lies almost exclusively in GPU power.
And you know this... how, exactly?

User avatar
Leto
Posts: 2031
Joined: Thu May 04, 2006 1:40 am
Location: Dune

Re: World #1 Go Player Ke Jie accepts Google Alpha Go Match.

Post by Leto » Tue May 23, 2017 5:24 pm

duncan wrote:
Laskos wrote:
Michel wrote:Only by 0.5 point if I understood correctly. On the other hand AlphaGo is programmed to go for the surest win and not the win that delivers the most points (which may involve taking higher risks). So the fact that AlphaGo "only" won by 0.5 points may be misleading.
0.5 without komi of 7.5? That was my impression, but I don't know how to count very well. From what I understood, Ke Jie tried to gain points, so that Alpha would need to attack, but mishandled the opening and in fact after the opening Alpha had more (potential) points and Ke had to attack. There was no opportunity to attack Alpha in the whole game, and Alpha, gaining advantage from the opening, just played quietly for the win. Alpha still seems to go to the highest probability, not highest advantage, so potentially it may had 10-15 stones advantage in late midgame, but went for the safer 6-8 points. I will maybe post Crazy Stone Deep Learning (7 amateusr dan) 2 hour analysis, that shows that from this engine (Zenith 6 too) point of view, White (Alpha) had substantial advantage right from the opening after move 20, a thing which seems to be overlooked or muted by 9p pros commenting. It seems that even relatively weak engines compared to those pros (3-4 stones difference) are better at scoring in the initial stages of the game.

Ke here seems to have had no any chances, a pretty overwhelming win by Alphago.
thanks for the analysis. keep it going!


with chess most (although not all) experts seem to be saying even with a 32 piece database, a computer with knight handicap would not be able to defeat a grandmaster.

do you have an opinion what handicap would be needed for a 9 dan go player to defeat a computer with access to a solved database of go.?
Are you sure about your knight odds comment? It seems Larry Kaufman said earlier this year that Komodo already plays about equal with a few grandmasters at knight odds in blitz. Here's where I found his statement:
http://talkchess.com/forum/viewtopic.ph ... night+odds

Seems to me that it's only a matter of time before engines can beat grandmasters at knight odds in long time control games.

User avatar
Laskos
Posts: 10181
Joined: Wed Jul 26, 2006 8:21 pm
Full name: Kai Laskos

Re: World #1 Go Player Ke Jie accepts Google Alpha Go Match.

Post by Laskos » Tue May 23, 2017 5:42 pm

Leto wrote:
duncan wrote:
Laskos wrote:
Michel wrote:Only by 0.5 point if I understood correctly. On the other hand AlphaGo is programmed to go for the surest win and not the win that delivers the most points (which may involve taking higher risks). So the fact that AlphaGo "only" won by 0.5 points may be misleading.
0.5 without komi of 7.5? That was my impression, but I don't know how to count very well. From what I understood, Ke Jie tried to gain points, so that Alpha would need to attack, but mishandled the opening and in fact after the opening Alpha had more (potential) points and Ke had to attack. There was no opportunity to attack Alpha in the whole game, and Alpha, gaining advantage from the opening, just played quietly for the win. Alpha still seems to go to the highest probability, not highest advantage, so potentially it may had 10-15 stones advantage in late midgame, but went for the safer 6-8 points. I will maybe post Crazy Stone Deep Learning (7 amateusr dan) 2 hour analysis, that shows that from this engine (Zenith 6 too) point of view, White (Alpha) had substantial advantage right from the opening after move 20, a thing which seems to be overlooked or muted by 9p pros commenting. It seems that even relatively weak engines compared to those pros (3-4 stones difference) are better at scoring in the initial stages of the game.

Ke here seems to have had no any chances, a pretty overwhelming win by Alphago.
thanks for the analysis. keep it going!


with chess most (although not all) experts seem to be saying even with a 32 piece database, a computer with knight handicap would not be able to defeat a grandmaster.

do you have an opinion what handicap would be needed for a 9 dan go player to defeat a computer with access to a solved database of go.?
Are you sure about your knight odds comment? It seems Larry Kaufman said earlier this year that Komodo already plays about equal with a few grandmasters at knight odds in blitz. Here's where I found his statement:
http://talkchess.com/forum/viewtopic.ph ... night+odds

Seems to me that it's only a matter of time before engines can beat grandmasters at knight odds in long time control games.
No. Level of Nakamura and better, Knight odds at LTC never.

Post Reply