AlphaGo’s Mastery Continues

Wuzhen, China. The last two days have seen Google DeepMind’s AlphaGo Master continue its declaration of superiority by defeating Ke Jie (9 dan professional) for a second time and gaining victory in a hard-fought battle against a whole team of five Chinese professionals, all of them 9 dan and representing the very pinnacle of ability.

Can nobody defeat Dr Hassabis’ robot?

One feature of AlphaGo’s play stood out above all: it is weirdly comprehendable.

I have seen more than a few professional game records and, without commentary, the fighting at the very top level often comes across as a kind of black magic but AlphaGo’s moves seem natural and rhythmical and their meaning is often immediately apparent even when they’re placed on points that no human player would find.

The next most unbelievable observation is that the engine plays some of our human joseki.

Instead of inventing new ideas, AlphaGo is simply showing us how to better employ ideas that have studied for hundreds of years. This is amazing because it shows us that the research of us mere mortals has been astoundingly accurate. Our romantic heuristics with which we reason and argue are somehow correct – sub-optimal they might be but they are not entirely wrong.

How is it possible that such a simple game played with primitive material and a mere handful of rules can provide a space in which such simple sequences can have so many layers of meaning and possibility?


I wonder what the future holds for AlphaGo and the game of Go.

We have seen AlphaGo discredit a lot of conventional wisdom that is taught like a mantra from the early kyu ranks, in the schools of the Insei and in professional study sesssions alike. To many, this opening of the mind has been a liberating experience comparable to the revolution in opening-theory pioneered before World-War II by Go Seigen and Kitani Minoru: the Shinfuseki.

“You have sente; you can play anywhere you like!”

I would like to ask AlphaGo many questions. I lost a game, once, to a rival who played what is known as the Great Wall opening – a crazy thing with a wavy line of stones down the middle of the board. I scoffed; I played high in all four corners and ended up a few points short of his total. What would AlphaGo think of such unconventional opening positions?

How strong would AlphaGo proove to be if you mandated that it starts at tengen – the point at the very centre of the go board?

Would AlphaGo find Honinbo Shusaku’s “ear-reddening” move?

How would AlphaGo handle a thousand-year ko?


I think that the game of Go, played between humans, will survive this onslaught by artificial intelligences simply because human play is sub-optimal. AlphaGo has shone a light on our imperfection and, honestly, it has been a painful experience but precisely this imperfection will be what saves the game.

The fact that a human opponent will never play perfectly, however strong they may be, allows opportunities for creativity. The fact that artistry governs a human’s opening and early game more than hard numbers or utility-scores allows a chance for exploration, experimentation and poetry.

If there is one thing to fear it is a future in which human professionals rely on bots too heavily, learning to mimick them and refusing to consider sequences that have not been ‘legitimised’ by some algorithm’s play. This is not to be feared because it will change the nature of competition. It will not lower the level of play between top players. It will merely be boring and bland.


Tomorrow, Ke Jie is to play a final match against the machine and he has politely requested to take the white stones – a request that DeepMind graciously granted. (Lee Sedol did the same in 2016.)

I have no doubt that tomorrow’s game will be spectacular but, despite the fact that both Ke Jie and AlphaGo appear to believe that playing white is advantageous, I think the best our human champion can hope for is a spirited performance and an honest loss in good style.

Advertisements

AlphaGo vs. Ke Jie: Game One

Yesterday, on the first day of the Future of Go Summit in Wuzhen, China, Google DeepMind’s AlphaGo (9 dan professional) neatly defeated Ke Jie (9 dan professional), the nineteen-year-old ‘final hope’ for human players of the ancient game. I watched the live-stream and saw the thing unfold.

I have been playing Go for many years and I am currently ranked in the middle of the “single-digit kyu” section of the amateur rank table in both Europe and South Africa. I watched, mused and wrote while last year’s Google DeepMind Challenge Match saw the upstart’s robot beat Lee Sedol (9 dan professional) – the Roger Federer of Go. In March 2016, this was a huge surprise to me as a computer scientist and Go player: it was too soon. It was a decade too soon!

I watched as Deep Zen – another of these new-generation of deep-learning-powered hybrids – failed to repeat DeepMind’s achievement when matched against Cho Chikun (9 dan professional) – he of the crazy hair, the tea drinker, almost definitely the most well-known Go player alive.

I read the stories and perused the records from the sixty games played online by the mysterious character who identified as Magist or Master and was later revealed to be none other than a new revision of DeepMind’s AlphaGo. The engine won fifty-nine of those games – all sixty if you grant her a victory for one game that ended prematurely due to technical troubles on the human-player’s end – but, by the end of the year 2016, this outcome was a foregone conclusion in my mind.

From all the evidence, it appears that AlphaGo has walked the path from first victory by an algorithm against human champion to last victory of a human champion against an algorithm in about one year. In the Chess world, this took a decade or more.

I might be wrong. Tomorrow, Ke Jie might strike back – I hope he does. It seems unlikely.


Ke Jie (9 dan professional) and AlphaGo’s handler – I think thought he was Fán Huī (2 dan professional) but he was actually Aja Huang (5 dan amateur) – took their places at the game-board and, with a few official statements by the coordinators and the formality of nigiri, the colours where assigned and the game began. Ke Jie placed the first, black stone at the three-four point: komoku.

The first surprise of the game was Ke Jie’s second move: the three-three point in the corner opposite to his first. The second was white’s creation of a two-space corner-enclosure with both stones high, on the fourth-line. Ke Jie invaded a white corner insanely early and white played a cut that seemed, at first, to be a little senseless. The game was certainly not boring in any way, although none of these anti-traditional plays appeared crazy to the eyes of the audience: amateur and professional players, alike, who had had their eyes and minds opened to such new possibilities by previous AlphaGo games.

The fashion in which AlphaGo calmly but inexorably drew ahead was reminiscent of last year’s event. There were no major upsets. Trades and exchanges of territory happened without drama but they are to be expected at such a high level. By the end-game, white’s advantage was so marked that she could afford a few inefficiencies for the sake of certainty and safety.

The result was another victory for the software: white won by half a point.


Numerically, half a point is the smallest margin for victory that the rules of the game permit and such an outcome is only possible for reasons of tradition and aesthetics: a draw would be considered undesirable and so white is granted the half-point as a tie-breaker. Despite all that, this was a very comfortable victory for the bot.

You could see evidence of the bot’s confidence of victory in the moves that were played during the latter stages of the game as well as the moves that were not. Ke Jie was the one who seemed desperate to claim super-optimal points while AlphaGo seemed content to capture stones that might present a later risk and to sacrifice a point here or there to ensure safety over all.

After the game, during the atrociously translated press conference, Ke Jie praised AlphaGo as a deity, stated his will to learn from the bot and, somewhat contradictorily, mentioned that this event will be the last time he plays against an artificial intelligence.

Ke vowed never again to subject himself to the ‘horrible experience’,” wrote The Guardian.

The Guardian and many other western news-sources painted Ke Jie as a sore loser, siezing his words and quoting them next to an unfortunate photograph, both out of context. They called him glum and, if one did not know better, one might simply swallow such a tale.

Go is a intense, immersive and deeply emotional game even at my middling amateur level. I can only believe that it is more so at the top. Horror and displeasure are emotions. Humans feel them. I am sure that all professionals have had moments of horror in many games – perhaps in all their games. Not for a moment will I accept that Ke Jie was complaining about his plight.

I think he was actually talking in general about the experience of playing against a machine – a stance that is further supported by the fact that he went on to say that he found games against other human beings to be more rewarding and more enjoyable.

I agree with Ke Jie on this count. Playing against another human player adds emotional, psychological and social dimensions that are simply absent from machines and are even somewhat lacking when playing across the Internet. Between two humans, Go is a sharing of mental creativity and an intimate communication that has nothing to do with words or language.

It is also less of a blow to lose to another living person than to a machine or a nameless avatar. Be they your sworn enemy, life-long rival or simply a personality who grates on your nerves, there is still some camaraderie, good-will and friendship.


I will watch tomorrow’s game, eagerly, and I am seriously intrigued by the pair-go and team matches that will follow on Friday and Saturday but, for me, the novelty of this style of contest is starting to wear thin.

I feel there is no artistry in the Go that is played by the algorithm. The beauty and poetry is missing. That it approaches optimality is not in question but optimality is somewhat boring.

When a master studies the record of a game he can tell at which point greed overtook the pupil, when he became tired, when he fell into stupidity, and when the maid came in with the tea.” – anonymous.

Robots neither drink tea nor require rest. They feel no emotion and no such impulse from their opponent will flavour their actions.

KGS Leben: On Hold!

Due to a career-related change of priorities occurring in the second half of this year, I have put KGS Leben on hold temporarily – I do not have time to work on it. (In all honesty, it has been quite a while since I have had time!)

For the time being, I recommend that interested Go players take a gander at Ilya Kirillov’s extremely exciting Go Universe project or dive into my KGS Leben source code and build whatever they want – I will gladly answer questions, merge pull-requests and patches and help when I have time.

KGS Leben: Autumn Update

Here, in the Northern hemisphere, Summer has rolled past and it is Autumn once again. The leaves are turning on the trees, the wind has picked up, the temperatures have plummeted, the weather in the Alps is cold and wet and Federweiße and every variety of Kürbis adorn the shelves in the German supermarkets.

I have finally found time to write gratuitous code and I even played a whole game of Go, yesterday!

I have committed to KGS Leben twice in the last week. Here follow the highlights.

Sound Cues

I was not too surprised by the clamour for clicks and beeps. I did not intend on implementing them, yet, but recording and integrating sounds seemed like a creative way to resume work on the project and they were clearly desired by players so here they are.

This click was recorded, manually, and retouched in Audacity to remove the noise of my desktop microphone, fade it in and out and clip it so that it feels synchronised with the board when in use.

The pass bell was synthesised using a script in Audacity that implemented the work of Jean-Claude Risset.

Both of these sound cues should be played in all supported browsers, now, although Internet Explorer will play them with the obsolete HTML 5 Audio object instead of the draft Web Audio API published by the W3C and implemented by nearly all other browsers of import. In that case, concurrent instances of the same sound cue will be culled – this might be noticeable in blitz games.

Unit Testing

I have added some unit tests for domain-models such as the game-position and game-tree models and I intend to add many, many more. I know I should have done this a long time ago because development is always smoother when one possesses ability to automatically test not only existing code but changes and new code too.

To write the unit tests, I have used Mocha.js. It works acceptably well, in general, and is actually quite good for JavaScript. In my development environment, I can even attach to the Node.js debugger and step through my code, using source-maps so that I can work with the original TypeScript sources. Development for the Browser is almost becoming civilised!

Microsoft Internet Explorer

The player-panels are no longer a complete dog’s breakfast in Microsoft’s quirky browser and several other odd glitches have been fixed. I have tested under MSIE 11 and under MS Edge and both appear to work.

The list of tested browsers now includes Firefox, Chrome, Internet Explorer 11 and Edge. Opera and Safari (desktop, not mobile) should be supported but I have not tested under them. Internet Explorer 10 might be usable – I will probably test it at some point, if only for amusement purposes.

Ko Centres marked on the Board

2016-10-01-ko-marker

The picture on the right says all that needs saying.

Behind the scenes, both the last-move marker and the Ko marker are now part of the game-position schema, making it incredibly easy to cover them in the unit tests for that model.

I also re-wrote the way that Ko is detected – the new implementation is much cheaper than the old one which relied on comparing positions in the active branch of the game tree.

The server implements the Ko-rule but it also necessary to implement it on the client in order to draw the marker and to avoid sending illegal moves to the server only to have them rejected.

Pass, Resign and Done Buttons

Prompted by feedback, I went searching for these three buttons and discovered that they were suffering from some sizing problems and they were completely hidden in Internet Explorer. These issues are now fixed. I have also tested them and they work as expected.

Here are the pass and resign buttons when the player-panels are positioned above and below the board:

This slideshow requires JavaScript.

When it is one’s opponent’s turn to move, the pass button is disabled. Clicking either button shows two secondary buttons. Clicking the second of these, the colourful one, will confirm the action, playing a pass-move or resigning the game. Clicking the other, or anywhere else for that matter, will cancel the action.

Here are the buttons when the player-panels are positioned on the left- and right-hand-sides of the board.

This slideshow requires JavaScript.

In this set of screenshots, you can also see the done button. During the scoring phase, the pass button morphs into the done button – a logical evolution.

KGS Leben is Open Source

Visit GitHub for the source code, along with instructions for building, testing and hosting the application. Features requests and bug reports may be submitted to the GitHub issue tracker for the project and open discussion can be found here, in the comments section, or in the Life in 19×19 thread related to the project. All project artefacts are released under the MIT License.

KGS Leben: Auto-Match

Earlier this afternoon, with my development environment still open and my code yet uncommitted, I connected to the live KGS server with one of my own ranked accounts and played the first real games ever to be played via KGS Leben. Games against other humans!

2016-05-31 hyaumet vs. Xharlie
hyaumet (5 kyu) vs. Xharlie (6 kyu, me)

I have reached a major milestone: all the use-cases required to play a complete game of Go (and those required to join the automatic match-making queue) are implemented, operational and stable enough for me to set aside my test accounts and play with a real one!

After playing several games, I joined a game between Imnot7d (7 dan) and Dom (6 dan) as a spectator and took some screenshots to show off the visible aspects of my work since my last post. I will upload a video showing the client in action as soon as I clean up a few rough edges that became apparent during today’s testing session but, for now, enjoy the pictures!

Firstly, by popular demand, I present markers on the most recently placed stone on the Go board:

Here is a picture of the board during the scoring phase:

Imnot7d (7 dan) vs. Dom (6 dan)
moku” markers showing territory and dead stones that were not removed during play during the scoring phase of a game between Imnot7d (7 dan) and Dom (6 dan)

Here is a picture of the score-summary section that appears in the sidebar:

Imnot7d (7 dan) vs. Dom (6 dan)
The score-summary table is updated dynamically during the scoring phase and also visible after the result has been finalised

The score-summary table draws some inspiration from the the dialogue box that the Java client, Cgoban, displays after the result has been finalised except, unlike the latter, my score-summary table is visible from the moment scoring begins and remains visible thereafter. I have also tabulated the data for readability and indulged in a little innovation. Let me state my case…

All the client software I have used to play Go on the Internet, when it does give the user a score break-down, adds the number of stones captured during play to the number left for dead on the board and marked as such during the scoring phase. A single tally of ‘captures‘ is presented to the user. This does not represent the way I count in my head.

When I count the game, mentally, I count territory and dead-stones together and add previously removed prisoners and constants such as komi or reverse-komi afterwards. For me, the final result of the game is important because it lets me know how accurately I did this and that feedback helps me get better at counting and, hopefully, stronger as a player.

In my score-summary, I have broken the player’s scores down to a more granular level, displaying separate counts for prisoners, being stones captured and removed from the board during play, and captures, being those remaining on the board at the end and marked as dead during scoring. I admit that the terminology needs some work but I think the feature will be very useful to players, like me, who want to know how close their mental counts came to the facts and where they went wrong in their assessment.

To the previously shown player-panels, I added buttons to pass and resign and, to the home view, buttons to join the auto-match queue. Joining incoming challenges has been doable for quite some time, as previously demonstrated and games started from incoming challenges can also be played to completion, naturally.

KGS Leben: Player Panels

I have finished working on the player-panels for KGS Leben, tweaking the automatic layout algorithm, polishing the seven-segment LCD clock’s design and adding elements to display capture counts, komi and user information.

Here are a couple of screenshots that I captured while spectating live games in the E.G.R.

2016-05-15 Player Panels - Away Landscape Canadian

In landscape orientation, the player-panels are displayed on the left and right sides of the game board. When spectating, the away panel (pictured above) belongs to the stronger of the two players, or the white player in the case that the players are of equal rank. When you are a player in the game, this panel will always belong to your opponent. Esquire55 (2 dan), playing white, must play twenty-two moves before the end of his rather short Canadian overtime period. White has not captured any of their opponent’s stones and is only receiving a half-point komi.

2016-05-15 Player Panels - Home Portrait Main

In portrait orientation, the player-panels are displayed as horizontal strips along the top and bottom edges of the board. This is a screenshot from a different game, showing the home panel. TOKKURI (2 dan), black, has captured two stones and has not yet entered overtime.

KGS Leben: Boards, Players & Clocks

I spent the last several days experimenting with KGS Leben‘s game area – the space that will hold the Go board, clocks, prisoner counts and more. At the start, I wasn’t sure where I was heading or where I would end up but something materialised, nevertheless, and I am immensely satisfied with it.

I divided the space into three areas: one for the board, one for the ‘away‘ player and one for the ‘home‘ player – thus named because, when spectating a game, neither of these may be the current user.

The two player-panels currently host the player’s clock and will show prisoner counts and other vital elements in the future. When one of the players is the current user, their panel will provide them with buttons including pass, resign and undo. Otherwise, the space will show the player’s identity: their name and avatar.

The layout adjusts according to the available screen real-estate, showing the away player’s panel above or to the left of the board and the home player’s panel below or to the right. The decision is made to maximise the size of the Goban, arguably the most important thing on the screen.

2016-05-05 Byo-Yomi Clock Feature
13×13 board at 2560×1440

The Goban itself is restricted in size so that it can never grow larger than a real-life specimen. This ensures that a nines board never grows so large that the stones appear like draughts-men! The user can always use their browser’s zoom-feature to bend this rule.

After settling on a design I implemented the game clocks. I began by using SVG graphics to make a simulacrum of my nice and chunky Chronos chess clock and, when that was operational, I began to extemporise…

This slideshow requires JavaScript.

A clear, readable game clock is crucially important and I think that this design satisfies that requirement with aplomb. Besides, if it looks this pretty after one quick iteration, it will be truly beautiful on release day!

KGS Leben Announced

KGS Leben is a prototype that will become a full-featured interface for the KGS Go Server. It is a web application and will simply run in a standards-compliant web browser, without the need to download or install any software or runtimes.

After one month of experimentation, exploration, trial, error, frustration, success, water, beer and tea, the software is capable of joining open game challenges, negotiating a game proposal with the challenger, playing moves on a Go board, spectating games being played by others, watching demonstration boards and chatting in rooms and channels. Furthermore, I have succeeded in the most difficult task demanded by all new software projects: I have chosen a name! The name comes from the German word for life „das Leben

None of the features exhibited in the preview video are really complete, yet, because I chose to spend the last month confronting risks that I identified when I began wrangling with the idea and tackling the most technically challenging facets of the project.

I hope to achieve rapid progress, now that the foundation has been cast.

KGS without Collateral Damage

The KGS Go server is almost certainly the most well-known Go server in the Western world, today. It is populated largely by friendly, English speaking players from American and European time-zones and, during popular hours, after work and on weekends, one is almost guaranteed a sufficiency of opponents at or very near to one’s own rank, from abject beginner to dan-level amateur. For beginners, it is a very good place to start playing online Go, for the more experienced, it is a good place to improve. Sadly, getting started on KGS is a little obtuse.

In light of the recent press frenzy surrounding the Google DeepMind Challenge Match and the great coverage that Go received, I present a tutorial in two parts that will dispel the fog of confusion that surrounds the fun and welcoming land of KGS. In this first part, I will explain how to install the KGS client, cgoban, without causing collateral damage to your Windows computer.

Official Method: Oracle’s JRE

If you have Oracle’s Java Runtime Environment installed or you are able and willing to download and install it, the official method is quite simple: navigate to the KGS Go Server homepage, download the executable file that will install cgoban, run it and follow the prompts.

If you cannot install Oracle’s JRE or, like me, you are repulsed by the thought of installing something which was bundled with malware in the form of the Ask Toolbar until very recently and now supposes to change your search provider and world-wide-web homepage, there is another option: you can use Zulu to run KGS and achieve an entirely portable installation with a very light-weight footprint.

Zulu is a build of OpenJDK, an open-source environment for Java programmes, and it can be downloaded as a primitive ZIP-file. Installation is a simple matter of extraction. It does not register processes that check for updates in the background, it will not nag you, constantly, to install said updates, it is not distributed with malware and, perhaps most importantly, it does not automatically integrate itself with your web browser — it does not present a security risk.

Note: at the time of writing, AZUL Systems only publish 64-bit builds of Zulu for Windows. My instructions, below, will not work for users on older processor architectures.

Portable Method: KGS with Zulu

Start by downloading Zulu from the AZUL Systems download page. I have tried several older versions, in the past, and version 8u72 at the time of writing — all of them worked.

2016-03-22 KGS without Collateral Damage 1

Each version has two download buttons: one for a ZIP file and one for an MSI installer. For this tutorial, you will want the ZIP format.

Next, download the KGS cgoban.jar file.

2016-03-22 KGS without Collateral Damage 2

You can do so by visiting the KGS home page and clicking the link entitled “Raw JAR File” or by navigating to: http://files.gokgs.com/javaBin/cgoban.jar

Once both downloads have completed, copy both cgoban.jar and the contents of the ZIP file into a new folder on your computer, henceforth referred to as your target folder.

2016-03-22 KGS without Collateral Damage 3

Create a new shortcut inside your target folder.

2016-03-22 KGS without Collateral Damage 4

The shortcut should point to javaw.exe — we will modify it to run cgoban in but a moment.

javaw.exe should be found in the ./zulu8.13.0.5-jdk8.0.72-win_x64/bin subdirectory of your target folder although your version numbers might be slightly different. You may name the shortcut anything you like — “KGS” makes a lot of sense.

After you have created the shortcut, right-click on it and open its properties window. Two changes must be made, as shown in the screen-shot, below.

2016-03-22 KGS without Collateral Damage 5

Firstly, append -jar cgoban.jar to the target of the shortcut. Ensure that there is a space separating this command-line switch from the existing text in the box and, if that existing text ends with a double-quote, append the switch outside the quotes.

Secondly, set the start-in path to your target folder which contains cgoban.jar. This can be done manually or by removing the last two levels (zulu8.13.0.5-jdk8.0.72-win_x64/bin) from the existing text. If the path contains any white-space characters, it will need to be enclosed in quotes.

After closing the properties window, double-click on your new shortcut to launch the KGS client.

2016-03-22 KGS without Collateral Damage 6

Because you downloaded Zulu from the Web, you are likely to be prompted with a security warning like the one shown above when you launch the programme for the first time. If your download was not corrupt and the executable’s digital security signature is intact, the publisher should be listed as ‘AZUL Systems Inc.’ and it is safe to uncheck the box captioned “Always ask before opening this file” and to click the run button.

2016-03-22 KGS without Collateral Damage 7

Congratulations — you have achieved an operational installation of CGoban 3, the client for the KGS Go Server.

Your shortcut to launch KGS can be copied to your desktop or pinned to Start or your task-bar just like any other shortcut in Windows. Before you do that, you might want to change its icon by using the change icon button on the shortcut’s properties window — I created the icon I use by cropping the banner image from the KGS server’s home page.

In the future, if you need to update the client, simply download the latest version of cgoban.jar and overwrite your existing copy. If you wish to uninstall the client, you only need to delete your target folder since everything resides therein. You can also copy your target folder to other computers although you may need to update the shortcut’s target and start in paths.

Have fun playing on KGS, comfortable in the knowledge that your computer has not been sullied. In the second part of this series, I will explain how to register a new user account on the server and talk about some issues that are not obvious to newcomers. (In light of my work on KGS Leben, my research for the second part of this article has ceased entirely. I had planned to talk about registering accounts, guest accounts, etiquette in the E.G.R., the escaper policy, the infamous tilde and other oddities. This may yet happen – one day.)

AlphaGo vs. Lee Se-dol: Haylee’s Synopsis

Hajin “Haylee” Lee, 3 dan professional Go player, secretary general of the International Go Federation, streamer, blogger, writer and creator of Haylee’s World of Go / Baduk, has published a long and somewhat rambling account of the recent challenge match between Google DeepMind’s AlphaGo (honorary 9 dan) and Lee Se-dol (9 dan professional) and I heartily recommend that you watch it.

It is a tale of clandestine phone calls and secret dinner parties, of people and the press and camera men and wives and daughters and friends. Haylee briefly touches on the games themselves, giving a light and digestible overview of the important moves in each game that will be immediately understood by anyone who plays Go and watched the games but her story is about more than the moves themselves — it offers a glimpse behind the scenes, told in a friendly and familiar and unassuming style by someone who was instrumental in organising the spectacle that millions watched, earlier this month.

At one point, during the video, Haylee singles out a camera man, named Greg, from Google’s documentary team, giving away clear evidence that a professionally produced story of the event is almost certainly in the works — something I would be very excited to watch.

Coincidentally, it was while I was perusing the text-chat during the American Go Association’s broadcast of the first game of the challenge match that I first discovered Haylee’s World of Go, Hajin Lee’s YouTube channel. I subscribed to it immediately and, since then, I have been enjoying her streamed games during which she plays Go on Tygem, a Korean server, and offers Malkovich-style comments on her thoughts and ideas and plans — her videos have become a regular source of entertainment and cerebral stimulation for me and, although I am only a single-digit kyu player, I find her commentary to be quite understandable and educational. I believe she has also published a book — an autobiographical account of her quest to become a professional player.

AlphaGo vs. Lee Se-dol: Game Five

Lee Se-dol resigned before the last handful of “yose” or end-game moves had been played in this morning’s game, bringing the Google DeepMind challenge match between AlphaGo, their Go-playing programme, and the 9 dan professional human player to a close. The artificial intelligence won the match by defeating Lee Se-dol in four of five games.

Lee Se-dol took the black stones and began with a solid, dignified opening that emphasised territory. He resisted white’s attempts to disrupt his plan and denied white the time to reinforce the boundaries of the areas of influence that white acquired in exchange. When a large group of white stones was cut off by a sequence that involved the infamous “stone tower” shape, it seemed like black was in the lead – he had gained thickness and resolved several weaknesses in exchange for a tiny number of points.

Slowly but surely, with calm and sensible late middle-game and end-game fighting, AlphaGo turned the game around once more and by the time the temperature of the end-game moves had fallen to nothing more than one point a move, it was apparent that black would be behind.

Lee Se-dol’s resignation was clear proof of his finely honed ability to count the score in his head because his deficit was barely measurable.

During the pre-match discussion, Google DeepMind talked about the fourth game of the match, the game that AlphaGo lost. They explained that move 78, the tesuji or skilful play that precipitated the demise of AlphaGo in that game, had “surprised” the engine and forced it to build a new plan for a branch in the game tree that it either hadn’t seen or had only explored perfunctorily. A human player in a similar situation would take a lot of time to re-evaluate and recalculate but AlphaGo’s time-control strategy is apparently very simple and primitive and it neglected to invest much of the abundance of time available on its clock.

Today, the post-match press conference was handled ineptly, the English translation coming and going and generally chaotic. A few interesting points could be gleaned from the chaos, including the fact that an ethics committee was set up inside DeepMind as a condition of their sale to Google.

Lee Se-dol also took ownership of his performance once again, saying that he believes that humans can do more against A.I. Go engines and admitting regret that he was unable to show us how. He reminded the audience of the importance of human creativity, said he began to question some classical beliefs about the game “a little bit”, and indicated that he had more studying to do.

AlphaGo’s skill with the stones is not superior to that of top human professionals, according to Lee Se-dol, its advantages are concentration and the psychological facets of the game.

During the ceremonies that followed the game, AlphaGo was awarded an honorary rank of 9 dan by the Korean Baduk Association.

This match is over but its culmination should be considered a beginning rather than an ending. Google DeepMind have executed a historic début, winning five-nil against Fan Hui (2 dan professional) towards the end of 2015 and four-one against Lee Se-dol, but AMD and Facebook and other contenders are yet on their way to the party. AlphaGo developed its strength through self-play – what will happen when these learning algorithms start to train against each other, like human insei, and new blood-lines are introduced to this gene pool?

In the coming months, we will learn more details about what happened in these five games. We will learn what Google DeepMind plan to do with their creation and how it will impact the worlds of Go, artificial intelligence and machine learning. We will see how Lee Se-dol moves forward in his own, human career. We live in exciting times.

AlphaGo vs. Lee Se-dol: Game Four

Lee Se-dol obliterated Google DeepMind’s AlphaGo with an inspired ‘wedge’ in this morning’s game, the fourth game of the challenge match between the 9 dan professional human player and the upstart A.I.

2016-03-13 __go4go_20160313_Google-AlphaGo_Lee-Sedol0000
The moves leading up to Lee Se-dol’s incredible move: white 78

Absolutely everyone showered praise on Lee Se-dol for his fantastic play, among them, Gu Li, a 9 dan professional player from China who is considered to be one of the strongest professional players on the planet and particularly relevant because of the international rivalry that exists between him and our champion facing DeepMind’s engine, today.

After the game, Lee Se-dol was finally ready to tell us where the weaknesses lie in this version 18 of the distributed AlphaGo programme. He said that the A.I. struggled, holding the black stones, and said that “surprises” like his skilful wedge in the centre forced “bugs” to show in the bot’s play.

He even put his money where his mouth is by explicitly requesting to play as black in Tuesday’s match. Presumably, now that he has identified a vulnerability, he intends to show the world that he can defeat AlphaGo even when it is playing its preferred role in the game.

During the press conference that followed the game, one reporter raised concerns that AlphaGo’s database of professional game records equipped it with extensive knowledge of Lee Se-dol but, to his disadvantage, the man knew almost nothing about the machine. He dubbed this imbalance “information asymmetry” and, in response to his question, Demis Hassabis, one of the founders of Google DeepMind, made some intriguing statements. Firstly, Mr Hassabis stated that AlphaGo’s training database did not contain any game records from professional games played by Lee Se-dol. He said that its database was populated with amateur dan-level games only and, from there, AlphaGo had trained by playing against itself. He pointed out that even a thousand records of real-world games would be insignificant amongst the millions of records created by self-play. What he says makes sense but I do wonder why they didn’t prime the system with professional game records in addition to amateur ones — it seems like an easy thing to do.

In the first three games of the challenge match, AlphaGo exhibited a high level of skill and mostly played moves that made sense, even to us humans. It earned the respect of its opponent and the Go community in general. Lee Se-dol said it was not unreasonable; I was among those who praised it for its human-like play. When it was ahead, it played moves that could be considered sub-optimal but even those were not absurd. Today, things were very different.

Even after Lee’s wedge, all was not lost, but AlphaGo sealed its fate by playing appallingly once behind on the board — many of its moves were simply ludicrous — and spurned its hard-won good-will by charging blindly onwards when its situation had become hopeless and the only responsible act was prompt and polite resignation. Both of these behaviours are familiar to anyone who experienced the laughable death-throes of the Monte-Carlo Tree-Search Go engines and, unfortunately, today’s performance presented an ugly glimpse of AlphaGo’s pedigree.

Yesterday, we saw proof that AlphaGo has conquered the mystical art of Ko and that assures me that, one day soon, these bots will also be able to fight to the fore after falling behind and learn to resign with dignity but, for now, DeepMind’s ‘prototype’ still needs work.

AlphaGo vs. Lee Se-dol: Game Three

This morning, Lee Se-dol lost a third and deciding game in the challenge match against Google DeepMind’s AlphaGo. The fourth and fifth games will be played but the prize goes to the programme and the prize money will be paid to various benevolent causes.

The game started with a high ‘Chinese’ fuseki which was soon broken up by a vital battle that affected the whole board, a product of Lee Se-dol’s aggressive style and AlphaGo’s refusal to capitulate. Black fought for his heavy and inefficient stones but white, played by the bot, acquired a huge territory that black simply couldn’t match. In desperation, black invaded and a scuffle lead to a complicated situation with multiple Ko fights. Black resigned when it became clear that his stones were lost and, elsewhere on the board, the cost of the scrap became too great to bear.

One incontrovertible fact was proven during the match: AlphaGo does not struggle with Ko fights in the slightest. AlphaGo showed that it can handle extremely tricky Ko situations that play havoc with the game-tree and would have completely flummoxed its ancestors, the MCTS engines of yester-year.

Lee Se-dol admitted that AlphaGo defeated him, taking ownership of his defeat and lamenting that he did not show us a better game, today. He said that today’s defeat was his defeat, not a defeat of human beings.

Michael Redmond, the 9 dan professional westerner who provided the official English commentary during the match, brazenly declared that AlphaGo “beat Lee Se-dol at his own game!” In front of a veritable army of reporters, in the ballroom of a hotel in down-town Seoul, he went on to suggest that AlphaGo would herald a “third revolution” in Go opening theory, mentioning AlphaGo in the same sentence as Honinbo Dosaku and Go Seigen, two legendary human players who sparked extensive, novel innovation in the fuseki in the past.

When a member of the audience questioned Redmond’s prophecy, Lee Se-dol answered that he thought AlphaGo is not at the level of the so called ‘Divine Gods’ and described the bot’s play as ‘different’ and ‘superior at times.’

Meanwhile, in the Go community, speculators questioned whether Lee Se-dol (9 dan professional) was, in fact, the best player to champion humanity in this duel against the machine. Some pointed to his aggressive and frequently precarious fighting style and hypothesised that AlphaGo’s cold and heartless play was ideally suited to winning in such situations – situations in which mere mortal, human professionals might become flustered or overwhelmed by the complexity. Some suggested that calmer, more placid players might fare better against the A.I.

Lee Se-dol reassured the audience at the post-match conference that AlphaGo is not yet perfect. I believe him. If he does not defeat AlphaGo, another will, but I do hope that he shows us the way over the course of the next two games.

AlphaGo vs. Lee Se-dol: Game Two

Lee Se-dol (9 dan pro.) conceded another match to AlphaGo, today, in Seoul. Google DeepMind’s lead now stands at two games to none and a straight-sets victory over the human player seems entirely possible. We will have to wait until Saturday to find out.

During the press conference that followed the game, Lee Se-dol admitted that he felt that black, played by AlphaGo, never fell behind or gave him an easy chance at any point during the match and that the programme made moves that he considered entirely reasonable. Reading between the lines, I suspect he even felt some sort of respect for the thing! He attributed his loss to an inability to find the bot’s weaknesses.

Google DeepMind claimed that the A.I. itself was confident of victory for most of the game, particularly during the opening and end-game phases, but that the team behind the scenes were nervous because the opinions of other strong human professionals were widely varied – some predicting that white, Lee Se-dol, would ultimately come out in front.

The development team also claimed that their only role, during the game, was ensuring that the programme had access to the computational resources and infrastructure that it required.

In the coming months, I would dearly like to know more about the level of human intervention that took place behind the scenes, the computational resources that powered the engine, and the learning and improvement that took place between the games of the match.

Personally, I will be cheering for the man who is opposing the machine, on Saturday. AlphaGo has already acquitted itself skilfully and Google DeepMind have already made a smashing début; a close match will be far more satisfying than a clean sweep by the engine.

While we wait to see what transpires in game three, it is time to start considering what AlphaGo’s victories mean for the future of the game of Go itself, the future of professional, competitive play and the future direction of research in the field of Computer Go. Perhaps we should all start reading and remembering the ripples that IBM’s DeepBlue sent through the world of Chess.

AlphaGo vs. Lee Se-dol

In the wee hours of this morning, I watched Google DeepMind’s AlphaGo programme defeat Lee Se-Dol (9 dan professional) in the first game of a five game match that will continue for the rest of this week. They played on a full-sized (19 by 19) board, AlphaGo played white and Lee Se-Dol paid komi. The field was entirely level, the game entirely fair. This simply wasn’t supposed to happen during my lifetime!

When I first encountered the game in 2007, my friend and fellow programmer promoted it as a game that bots simply couldn’t play. At that time, GNU Go was available but barely stronger than an improving double-digit kyu journeyman like myself – I soon learned to defeat it. A few bots exploited Monte Carlo Tree Search (MCTS) and they could achieve low single-digit kyu strength on a regular desktop computer or amateur dan strength on exceptional hardware but that was about all.

Everyone knew that MCTS was the way forward, that heuristics used in the random Monte Carlo simulations would improve with time, and that the inexorable increase in exploitable computational power would permit an ever-increasing number of simulations to be run. Everyone knew that even the professionals would fall, one day. That day was supposed to be decades away and the DeepMind team proved everyone wrong.

This morning, the outcome of the game was far from the only impressive feat that DeepMind’s programme demonstrated: the human-like nature of its moves and the mature management of the clock were a wonder to behold, as was the calm and steadfast way in which it handled Lee Se-dol’s cunning but unorthodox opening that seemed like an attempt to deviate from more common and well-known openings that would exist in databases and opening books that the artificial intelligence was surely using.

I was strangely sad when I saw Lee throw in a prisoner to indicate his resignation, this morning. In that instant, something ineffable changed in a way that can never be reversed. I will continue to watch in fascination as this match unfolds but, even if Lee wins the four remaining games, history was made this morning and I am very happy to have watched it as it happened.