The Future of Go Summit: Pair Go moves analysis

Google DeepMind · Advanced ·📄 Research Papers Explained ·9y ago

Key Takeaways

Analyzes Pair Go moves with commentary from DeepMind research scientist Thore Graepel and former professional Go player Hajin Lee

Full Transcript

hello from the future of go Summit in vuan China I'm Tor grapel research scientist at Deep Mind I'm joined here by hajin Lee a former forun professional player hello so we had a very special format today for the first game can you tell us about that yes the first game was Parago format where two human players each of them teamed up with alphaco and then they really had to work as a team but they were not allowed to discuss with each other oh fascinating and how did that work out so it was really interesting to see how the human players were like trying to adjust to the alphago style and in that process I believe the players learned a lot of different things oh fascinating so uh let's look at an early position what happened here in the top right corner that's right so in this game when black came over here we all expected white to push over here because the normal sequence and black would extend white would jump but then this ji over here wouldn't really coordinate with this part because when black pushes white need to answer this way and black can even push it more and with this shape this white would be somewhat over concentrated so when black comes over here this result would be FL uh favorable for black right so uh White uh played a different move then right that's right so at this point it was alha Go's turn in White's team and ala go to this one space jump which is often considered um bad move that's because when black pushes here and then presses white has to come out and then when black Cuts this is a Mii situation where white has to give up either this two stones or this one stone right but in this particular situation it worked out remarkably well can you show us how yes that's right so this situation was special because after black captur this one stone this Black's um good influence was somehow reduced by this White's position already and also because this was leather white could get two moves around here so when white came over here if black Ed the then white could save this one stone right so that's why black had to come over here and we can see that still these white stones are somehow limiting the black's influence toward this area and then white could come over here and make a second attack yeah that's right fascinating yes thank you okay so here a little later in the same game um another interesting situation arose can you explain what happened that's right so after black came over here white plate here and this move over here is looking at Black's weakness so if black answered here to reinforce this weakness white would extend here making a good shape in the side but it is somewhat unavoidable for black because this um attach over here would be too severe yes in the actual game black played here which was uh amazing timing in my opinion because often times when you see this Invasion white needs to let this black connect to the corner in order to keep this white shape but if white did that in this situation by connecting to the corner black would this Black's weakness over here would be naturally weakened so for example after this one because the corner is already quite safe when white needs to come out over here to secure this group black could choose this one and then this one wouldn't have much meaning for white so white would have to choose a different variation here that's right so that's why white didn't want to let black connect to this corner in the actual game white exchanged this attach and then H it so that if black blocked here white would block this one and then when black Ataris instead of connecting here white would choose this cut and after this black can save this group right but there is still this weakness and then also white can have a better shape in the center so this was White's plan to to like this right but that's also not what happened then that's right yes and because White had this plan instead of blocking this corner black chose to come out here and then white exchange it this one if black blocked here then it would become the same sequence right but black connected here and then white Tar this one So eventually it became a big exchange where white got this corner and then black got the outside so HED here connect it would have been nice if white could come out here but then the capturing race between the corner and the outside would have been some difficult for white right yes so like this it would have been not possible for white so white had to come back to the corner here and then black blocked White had to connect and then black blocked and then black blocked here at the post game press conference fui told us that IO thought this result was good for black right yes because although white got the whole Corner Black's outside was really good and also black got sente and this one was sent so white had to take this one and then with here right and so white got the corner territory but black ended up with a lot of strength on the outside that's right yes thank you so later in the game there was an interesting development here on the lower left can you explain what happened that's right so after black built this influence came into here to attack these two stones and then we all thought this was a good move right but at the press conference fanu told us that actually Alpo was expecting this move and the merits of this move would be yes and this move is very difficult to think because first of all it's very high on the fifth line and also when you think about um chasing these two stones somehow eliminating the eye space is the first intuition you have but this move is um globally looking at all this um blacks influence the whole board and somehow because of this uh sent moves white cannot come outside so even if white could live inside black would be able to build a huge moo in the center and that would give blacks big Advantage so that was much better move than this Invasion and when Gully saw this move for the first time he said that he wouldn't have thought about this move once he saw it he thought it was quite impressive yeah it's very good to see that professionals can learn from alphao that's right yes great thank you thank you

Original Description

Catch up on the most exciting moves from Pair Go at The Future of Go Summit. With commentary from DeepMind research scientist, Thore Graepel, and former professional Go player and Secretary General of the International Go Federation, Hajin Lee.
Watch on YouTube ↗ (saves to browser)
Sign in to unlock AI tutor explanation · ⚡30

Playlist

Uploads from Google DeepMind · Google DeepMind · 38 of 60

1 RL Course by David Silver - Lecture 8: Integrating Learning and Planning
RL Course by David Silver - Lecture 8: Integrating Learning and Planning
Google DeepMind
2 RL Course by David Silver - Lecture 1: Introduction to Reinforcement Learning
RL Course by David Silver - Lecture 1: Introduction to Reinforcement Learning
Google DeepMind
3 RL Course by David Silver - Lecture 2: Markov Decision Process
RL Course by David Silver - Lecture 2: Markov Decision Process
Google DeepMind
4 RL Course by David Silver - Lecture 5: Model Free Control
RL Course by David Silver - Lecture 5: Model Free Control
Google DeepMind
5 RL Course by David Silver - Lecture 6: Value Function Approximation
RL Course by David Silver - Lecture 6: Value Function Approximation
Google DeepMind
6 RL Course by David Silver - Lecture 4: Model-Free Prediction
RL Course by David Silver - Lecture 4: Model-Free Prediction
Google DeepMind
7 RL Course by David Silver - Lecture 3: Planning by Dynamic Programming
RL Course by David Silver - Lecture 3: Planning by Dynamic Programming
Google DeepMind
8 RL Course by David Silver - Lecture 10: Classic Games
RL Course by David Silver - Lecture 10: Classic Games
Google DeepMind
9 RL Course by David Silver - Lecture 7: Policy Gradient Methods
RL Course by David Silver - Lecture 7: Policy Gradient Methods
Google DeepMind
10 Google DeepMind: Ground-breaking AlphaGo masters the game of Go
Google DeepMind: Ground-breaking AlphaGo masters the game of Go
Google DeepMind
11 Match 1 - Google DeepMind Challenge Match: Lee Sedol vs AlphaGo
Match 1 - Google DeepMind Challenge Match: Lee Sedol vs AlphaGo
Google DeepMind
12 Match 2 - Google DeepMind Challenge Match: Lee Sedol vs AlphaGo
Match 2 - Google DeepMind Challenge Match: Lee Sedol vs AlphaGo
Google DeepMind
13 Match 1 15 min Summary - Google DeepMind Challenge Match
Match 1 15 min Summary - Google DeepMind Challenge Match
Google DeepMind
14 Match 3 - Google DeepMind Challenge Match: Lee Sedol vs AlphaGo
Match 3 - Google DeepMind Challenge Match: Lee Sedol vs AlphaGo
Google DeepMind
15 Match 2 15 Minute Summary - Google DeepMind Challenge Match 2016
Match 2 15 Minute Summary - Google DeepMind Challenge Match 2016
Google DeepMind
16 Match 3 15 Minute Summary - Google DeepMind Challenge Match 2016
Match 3 15 Minute Summary - Google DeepMind Challenge Match 2016
Google DeepMind
17 Match 4 - Google DeepMind Challenge Match: Lee Sedol vs AlphaGo
Match 4 - Google DeepMind Challenge Match: Lee Sedol vs AlphaGo
Google DeepMind
18 Match 4 15 Minute Summary - Google DeepMind Challenge Match 2016
Match 4 15 Minute Summary - Google DeepMind Challenge Match 2016
Google DeepMind
19 Match 5 - Google DeepMind Challenge Match: Lee Sedol vs AlphaGo
Match 5 - Google DeepMind Challenge Match: Lee Sedol vs AlphaGo
Google DeepMind
20 Match 5 15 Minute Summary - Google DeepMind Challenge Match 2016
Match 5 15 Minute Summary - Google DeepMind Challenge Match 2016
Google DeepMind
21 DQN SPACE INVADERS
DQN SPACE INVADERS
Google DeepMind
22 DQN Breakout
DQN Breakout
Google DeepMind
23 Asynchronous Methods for Deep Reinforcement Learning: Labyrinth
Asynchronous Methods for Deep Reinforcement Learning: Labyrinth
Google DeepMind
24 Asynchronous Methods for Deep Reinforcement Learning: MuJoCo
Asynchronous Methods for Deep Reinforcement Learning: MuJoCo
Google DeepMind
25 Asynchronous Methods for Deep Reinforcement Learning: TORCS
Asynchronous Methods for Deep Reinforcement Learning: TORCS
Google DeepMind
26 Differentiable neural computer family tree inference task
Differentiable neural computer family tree inference task
Google DeepMind
27 StarCraft II DeepMind feature layer API
StarCraft II DeepMind feature layer API
Google DeepMind
28 DeepMind Health – Partnership with the Royal Free London NHS Foundation Trust
DeepMind Health – Partnership with the Royal Free London NHS Foundation Trust
Google DeepMind
29 DeepMind Health – Michael Wise – a patient's journey
DeepMind Health – Michael Wise – a patient's journey
Google DeepMind
30 Streams – a platform for a digital NHS
Streams – a platform for a digital NHS
Google DeepMind
31 DeepMind Lab - Nav Maze Level 1
DeepMind Lab - Nav Maze Level 1
Google DeepMind
32 DeepMind Lab - Stairway to Melon Level
DeepMind Lab - Stairway to Melon Level
Google DeepMind
33 DeepMind Lab - Laser Tag Space Bounce Level (Hard)
DeepMind Lab - Laser Tag Space Bounce Level (Hard)
Google DeepMind
34 Exploring the mysteries of Go with AlphaGo and China's top players
Exploring the mysteries of Go with AlphaGo and China's top players
Google DeepMind
35 Demis Hassabis on AlphaGo: its legacy and the 'Future of Go Summit' in Wuzhen, China
Demis Hassabis on AlphaGo: its legacy and the 'Future of Go Summit' in Wuzhen, China
Google DeepMind
36 The Future of Go Summit: AlphaGo & Ke Jie match 1 moves analysis
The Future of Go Summit: AlphaGo & Ke Jie match 1 moves analysis
Google DeepMind
37 The Future of Go Summit: AlphaGo & Ke Jie match 2 moves analysis
The Future of Go Summit: AlphaGo & Ke Jie match 2 moves analysis
Google DeepMind
The Future of Go Summit: Pair Go moves analysis
The Future of Go Summit: Pair Go moves analysis
Google DeepMind
39 The Future of Go Summit: AlphaGo & Ke Jie match 3 moves analysis
The Future of Go Summit: AlphaGo & Ke Jie match 3 moves analysis
Google DeepMind
40 Emergence of Locomotion Behaviours in Rich Environments
Emergence of Locomotion Behaviours in Rich Environments
Google DeepMind
41 StarCraft II 'mini games' for AI research
StarCraft II 'mini games' for AI research
Google DeepMind
42 Trained and untrained agents play StarCraft II full 1vs1 game
Trained and untrained agents play StarCraft II full 1vs1 game
Google DeepMind
43 DeepMind open source PySC2 toolset for Starcraft II
DeepMind open source PySC2 toolset for Starcraft II
Google DeepMind
44 ICML 2017: Test of Time Award (Sylvain Gelly & David Silver)
ICML 2017: Test of Time Award (Sylvain Gelly & David Silver)
Google DeepMind
45 Ke Jie and DeepMind's Go Ambassador Fan Hui review the 3rd AlphaGo vs Ke Jie game
Ke Jie and DeepMind's Go Ambassador Fan Hui review the 3rd AlphaGo vs Ke Jie game
Google DeepMind
46 Ke Jie and DeepMind's Go Ambassador Fan Hui review the 1st AlphaGo vs Ke Jie game
Ke Jie and DeepMind's Go Ambassador Fan Hui review the 1st AlphaGo vs Ke Jie game
Google DeepMind
47 Ke Jie and DeepMind's Go Ambassador Fan Hui review the 2nd AlphaGo vs Ke Jie game
Ke Jie and DeepMind's Go Ambassador Fan Hui review the 2nd AlphaGo vs Ke Jie game
Google DeepMind
48 AlphaGo Zero: Discovering new knowledge
AlphaGo Zero: Discovering new knowledge
Google DeepMind
49 AlphaGo Zero: Starting from scratch
AlphaGo Zero: Starting from scratch
Google DeepMind
50 Defining principles for tech companies in the NHS: DeepMind Health's Collaborative Listening Summit
Defining principles for tech companies in the NHS: DeepMind Health's Collaborative Listening Summit
Google DeepMind
51 A systems neuroscience approach to building AGI - Demis Hassabis, Singularity Summit 2010
A systems neuroscience approach to building AGI - Demis Hassabis, Singularity Summit 2010
Google DeepMind
52 Retour de Rémi Munos en France et ouverture de DeepMind Paris
Retour de Rémi Munos en France et ouverture de DeepMind Paris
Google DeepMind
53 Grid cells - Caswell Barry, UCL
Grid cells - Caswell Barry, UCL
Google DeepMind
54 DeepMind Health Research and Moorfields Eye Hospital NHS Foundation Trust: What our research shows
DeepMind Health Research and Moorfields Eye Hospital NHS Foundation Trust: What our research shows
Google DeepMind
55 DeepMind Health Research and Moorfields Eye Hospital NHS Foundation Trust: A Patient's Story
DeepMind Health Research and Moorfields Eye Hospital NHS Foundation Trust: A Patient's Story
Google DeepMind
56 Deep Learning 3: Neural Networks Foundations
Deep Learning 3: Neural Networks Foundations
Google DeepMind
57 Deep Learning 5: Optimization for Machine Learning
Deep Learning 5: Optimization for Machine Learning
Google DeepMind
58 Deep Learning 8: Unsupervised learning and generative models
Deep Learning 8: Unsupervised learning and generative models
Google DeepMind
59 Reinforcement Learning 1: Introduction to Reinforcement Learning
Reinforcement Learning 1: Introduction to Reinforcement Learning
Google DeepMind
60 Deep Learning 2: Introduction to TensorFlow
Deep Learning 2: Introduction to TensorFlow
Google DeepMind

Related Reads

Up next
Welcome to the Next Temperamental Era
Charles Schwab
Watch →