Does this mean that a computer will soon be beating people at Go?
Japanese original http://www.nihonkiin.or.jp/event/area/other/post_704.html
Google translation https://translate.google.com/translate?hl=en&ie=UTF8&prev=_t...
The translation is pretty bad on technical terms. Obviously it's not about chess or shogi, but about go, and the "ten stage" should be the Tengen title, currently held by Ida Atshushi, the human player. I think this is the first time a Japanese title holder plays in one of these matches. It will be a 4 stone match. If the computer wins it should be at least 5 dan, maybe 6 dan (amateur scale). One should be at least 7 dan to be as strong as the weakest pros.
1dan -> 9dan (amateur)
1dan -> 9dan (pro abbrev. to 1p -> 9p)
I think 8d ama can start to have competitive games with 1p or 2p... I mean they're getting into the same ball-park in terms of strength but averaged out over time a rank above _will_ beat a rank below, that's why they are in the ranks they are in after all :)
See: https://matthieuw.github.io/go-games-number/GoGamesNumber.pd...
At the beginning level, players must learn and memorize a large number of Dingshi, which means fixed patterns. Their scopes are small, restricted to less than 30 cross points, if my memory serves right. They mostly happen in the corners or close to the borders, which helps greatly reduce the complexity. If one masters these Dingshi, s/he can easily beat ~90% players based on my experience. This part of the game can be done easily by computers.
The next level is to develop skills or feels to identify Xing, which refers to larger patterns. They are still constrained to a small area, like 1/8 ~ 1/6 of the whole board. They are further away from the corners and the boarders. So the complexity increases dramatically. Most players of intermediate level can feel if a Xing is good or not, but not many of them knows why. So s/he develops the intuition, but not the actual skills and insights. That means, if two Xings can be developed from the current state and both feel good, they can't tell which is better in most cases. In reality, there could be dozens of good Xings developed from the current state, and one also needs to consider the potential moves of the opponents. Due to the complexity, the vast majority of players can't go beyond this level no matter how hard they work on. It is like it requires a combination of talent, personality and hard work to become a top mathematician or a NBA starter.
The next level requires one's skills and insights to understand Shi, which includes strategy of the whole game, potential, etc. It is to develops the skills and feeling on the whole board. It is to develop accurate calculation on the potential covering area of the existing stones and how strong the potential is. It asks the questions that how synergetic are two regions of your stones , although they may look far apart at present. At this level, players pay great attention to the beginning stage of the game. The board is vastly empty, so the complexity is the highest. The potential gains or losses can be huge. One must consider the opponent's playing history to take fully advantage. It happens often to the players who enter the early stage of this level that they realized in the later stages of the game that one stone should have been placed one position left or right or higher. This initial stage is called Buju.
I think computers can get to the second level but it looks very hard to get to the third level in the next 10 years. Keep in mind that there are big differences in term of one's skills and insights even within the highest levels. For amateur players and low level professionals, they are way beyond their understanding. Even it is explained a move would be better if placed just one spot left, they wouldn't understand it.
In short: use Bayesian logic against the following question: "Calculate the probability of winning given that I've moved a stone at X location". Use monte-carlo to attempt to estimate hard numbers.
I'm not familiar with your terminology... my school of Go uses the term "Tesuji" (Japanese) to describe what you seem to call "Dingshi" (sounds like a Chinese name to me...). In any case, a computer with strong Tesuji (or maybe Dingshi in your terminology), combined with a monte-carlo method to look for "long-term strategic" moves is what has gotten Go Computers to where they are today.
Humans are superior at parallel processing. Go players activate the brain region of vision, and literally think by seeing the board state. A lot of Go study is seeing patterns and shapes... 4-point bend is life, or Ko in the corner, Crane Nest, Tiger Mouth, the Ladder... etc. etc.
Go has probably been so hard for computers to "solve" not because Go is "harder" than Chess (it is... but I don't think that's the primary reason), but instead because humans brains are innately wired to be better at Go than at Chess. The vision-area of the human's brain is very large, and "hacking" the vision center of the brain to make it think about Go is very effective.
Xing is likely 行 or 形, either having the meaning-sense of "shape" and "motion".
Shi is 勢, a very old word used in many different fields. It is used in the Chinese military lingo describe the structure and posturing of the battlefield, and Sun Tzu devoted a whole chapter of the Art of War to discuss it. It's used in Chinese martial art to describe a posture and all of it's potential and projections of power, similar to the Japanese martial art term "kamae". Considering that the generals and leaders of ancient China also used Weiqi to hone their strategic thinking skill, not surprising 勢 is used to refer to whole-board vision.