The first game between Ke Jie and AlphaGo at the Future of Go Summit was filled with suspense and intrigue, with nuanced, precise play from both Black and White. After many lively exchanges and innovative variations, AlphaGo, playing White, held strong and won by 0.5 points.
Ke Jie opened the game with a bold declaration by employing two of AlphaGo's favorite moves. The first of these, the 3-3 corner enclosure introduced to opening theory by the legendary players Go Seigen and Kitani Minoru in the 1930s, saw frequent use for decades before falling out of favor in contemporary Go. However, it is now seeing a revival as Ke Jie and other players have begun experimenting with it in recent tournament games. After AlphaGo responded by securing its own corner with its favored two-space high extension, Ke Jie leapt into the upper-left corner with a 3-3 invasion. Previously dismissed as unplayable, this move has experienced a sudden surge in popularity after AlphaGo introduced several new variations in its recent series of 60 games.
A rare variation of the 3-3 invasion joseki followed, at the end of which AlphaGo played a very interesting novelty: instead of resolving the ladder with a direct capture or the usual knight's move, it played on a grand scale with the large knight's move of 24. Fan Hui believes that AlphaGo was telling us its own unique philosophy: "AlphaGo's way is not to make territory here or there, but to place every stone in a position where it will be most useful. This is the true theory of Go: not 'what do I want to build?', but rather 'how can I use every stone to its full potential?'"
An exciting interchange on the upper left side followed and Ke Jie came through in excellent form. Giving up the corner for the side, Black captured four of White's stones in a trade that AlphaGo evaluated as ideal for both players. Not for nothing is Ke Jie regarded as the best player in the world! With the approach at 49, Black challenged White's influence on the lower side, but White turned lightly aside with the peep and cut at 50 and 54. These moves aimed not at a direct follow-up, but at subtly maximizing the potential and future initiative in the area. Although AlphaGo would have preferred the one-space jump to emphasize the center, Ke Jie responded by tightly gripping White's four stones at 51, maintaining maximum local control. With 55, a sharp probe that soon spurred Black to trade the lower side for the lower-left corner, Ke Jie made a choice that set the direction for the rest of the game. After Black lived in the corner, giving White an iron wall outside in compensation, Ke Jie resolutely discarded his stones on the lower side to gain a strong shape and initiative to return to the top.
Now behind in the balance of territory, Ke Jie was obliged to make the most of the upper side, and attempted to do so on a grand scale with the ambitious two-space jump of 97. AlphaGo's collected response at 98 forced Ke Jie to cut off the single white stone with 99, a decisive move that initiated the last major exchange of the game. Heading into the endgame, Ke Jie responded with vigour, but AlphaGo emerged with a modest but secure lead, ultimately winning by a half point.
We hope the innovations we have seen today are only the first of many, and look forward to Go players around the world analysing these moves as well as trying them out in their own games.