Go Legend Chapter 140

Chapter 140 Buddha is also angry

The "first dog trick" that came out of the "generation dog", why is "vulgar thorn" not the key, and the next step is the essence? To understand this problem, Li Xiangping believes that it is best to connect it with the original human tactics, and use the "dog strokes" to compare with human stereotypes, and then maybe we can see it more clearly.

The one that appears on the chessboard today is called "Xiaomu Tuo Retreat", which is one of the most basic patterns in Go. It is a very common, very simple, but also very old and very classic pattern.

It is said that this formula is very old, because the Japanese gave up the seat to play "Xiaomu Go", this formula has already appeared, and said that this formula is very simple, because the number of moves is small and the chess structure is simple, which does not include No matter how complicated the calculation is, there is no "flying knife".

In addition to these characteristics, there are few variations of this formula. From the day it was born to the present, it has remained almost the same. This was the case for Dao Ce and Xiu Ce a few hundred years ago, and after the 21st century, everyone Still like this. It is precisely because this is such a classic formula that Li Xiangping may be able to better understand the difference between human thinking and Go AI thinking by analyzing the difference between it and "dog move".

Anyone who understands Go knows that when the Go AI moves the "vulgar thorn", the human action is not like this. At this time, the normal progress of the "Xiaomutuo retreat" is to "remove the edge" in the next step, and wait until After Black has played this "edge removal", the entire local formula is completed.

As for the reason for this step of "removing the edge", it is of course well understood according to the original human Go theory. The three pieces of black that were originally made into the "tiger's mouth" have not been completely stabilized at this time. Three, three and four", so at this time, it is necessary to "remove the edge" in the next step. This move not only makes the black piece more stable, but also has a certain number of points.

Incidentally, it is none other than Li Xiangping's external chess master Shi Daqi who summed up the classic Go proverb of "standing two, splitting three and splitting four". No one has ever objected, so naturally, no one has ever objected to this step of "removing the edge" of the "Xiaomu Tuo Retreat", and everyone thinks that this is the most normal way to proceed.

And all this, until the "Alpha Dog" turned out, the situation changed!

So is the "dog move" overturning the Go proverb summed up by humans? From a strict point of view, this cannot be said. Even Li Xiangping, who has seen the power of Go AI, does not dare to draw this conclusion lightly.

However, as far as this "small eye support retreat" is concerned, if the "vulgar thorn" of the Alpha dog is compared with the "edge removal" of human beings, then Li Xiangping still can't do it. If you don't admit it, maybe it's really a dog move that seems more reasonable. This move is in line with a higher level of chess theory.

To put it another way, even until now, Li Xiangping doesn't think that the move of "edge removal" made by humans is a bad move, but compared to the dog move, this move seems too ordinary, and it is the so-called mediocre move. of it.

This is a bit like nitpicking. Now let's compare these two moves with the harshest eyes and see how mediocre this “edge removal” move is.

The first point: Humans make this move to "remove the edge", then after the completion of this formula, black will lose a second move, and Alphago's "vulgar stab" move and then take the first move, black will earn a first move locally. .

The gap between the first and the second moves in Go is definitely not necessary for chess players who understand chess. Even many people who don't understand chess should know that the "first move" in Go is better than the "second move", so this is the first difference. Needless to say, the dog who can earn the first move gets one point first.

The second point: The step of "removing the edge" under the human formula is actually not very valuable at this time.

Everyone should know that when this "small eye support retreat" appeared on the chessboard, it was not long after the overall situation began. There were only more than 10 moves in total, and the chessboard was still very empty. Big field", but by no means the biggest "big field" on the chessboard.

For example, in the chess game between AlphaGo and Li Shishi, as well as in the actual game of today's chess game, after black chess has taken the lead, he immediately goes to grab another "big game" on the chessboard, that is, let black chess form a "Chinese flow" on the other side. " formation.

There is no doubt that only from the perspective of the "big field", the value of the move that constitutes the "Chinese flow" formation is obviously greater than this move of "edge removal". Therefore, this second round should also be a dog move to score.

The third difference between the two methods, Li Xiangping believes that this is actually the most important difference:

Everyone knows that there is a saying in Go that "urgency" is greater than "big field", and the so-called "urgent" is actually a point of contention between the two sides. , such chess can be called "urgent". Another example is that a certain move is related to the life and death of yourself or the opponent. If you don't make this move, then your own block is likely to die, or if you don't make this move, you can kill the opponent's piece, but you can't. , then this kind of chess is called "urgency."

Then the question now comes, can the move of "edge removal" in the human formula be called "emergency"?

The answer is obvious, of course, this move is not a 'emergency'. Therefore, the direction of this step of "removing the edge" is the saline-alkali land area commonly known in Go, and playing chess in the direction of the saline-alkali land is of course valuable, but it certainly does not involve the mutual elimination of the two sides' battles. Therefore, from this perspective, Of course, this move can hardly be called "emergency".

Second angle: If Black does not make this move to "remove the edge", will the large piece of Black die? Take a step back and say that even if you don't die, will you be miserable by White?

Yes, Li Xiangping thinks this is the crux of the problem!

This is because the original intention of human beings to 'remove the edge' is to take this move as an 'emergency', thinking that if they don't make a move here, black may be severely attacked, so it is necessary to reinforce one move. It is precisely because this move not only strengthens itself, but also has a certain practical value, so for thousands of years, very few people have raised objections to it.

It's a pity that AlphaGo thinks this move is not "urgent"!

And after seeing that dog move, after careful research and a lot of actual combat verification, humans must also admit that the step of "removing the edge" is indeed not a rush.

The reason is very simple. Black's original "tiger mouth" chess form is actually very flexible, especially after adding the "vulgar thorn" move, the eye position is actually very rich. From a safety point of view, the effect may be better than The human's move of "breaking the edge" is almost, but it is absolutely impossible for White to attack the ground badly, and it is even less likely that there is a danger of life and death.

And this is actually the real connotation of the "first dog move" in this step!

For this dog move, first use a 'vulgar move' to reinforce oneself. Although this move may not be as safe as the human's move of "edge removal", considering the first two advantages of this move, then In a comprehensive comparison, it is of course embarrassing to think that this step is more reasonable, and the human step of "removing the edge" is set off as a mediocre work.

Is the connotation of this "dog move" very deep? Li Xiangping believes that it is not really advanced, because such a move is only played by the "generation of dogs" after all, and everyone knows that the generation of dogs that use the "deep learning" algorithm is still learning how to play by humans, so Its Go thinking is actually more or less human.

For example, I have analyzed so much just now. In fact, in a word, this "first dog move" of AlphaGo is commonly known as "getting the first move".

The biggest feature of this kind of play is actually to give up part of the security within the range that can be accepted by oneself, and then emphasize the rapid development of chess and quickly seize other big games.

There are also practitioners of in human chess players, and the most typical example is of course Mr. Wu Qingyuan.

Li Xiangping remembers that in real history, when McClemon 9 dan was explaining the game between Li Shishi and AlphaGo to European and American people, he emphasized it many times; he clearly saw Mr. Wu's shadow in AlphaGo's chess game. , such as the "five-way sharp rush" that shocked everyone's attention later, this move is very much like Mr. Wu's chess.

Li Xiangping believes that it is not only that move, but the seemingly unbearable "thorn" of this move is actually in the same vein as Mr. Wu's Go thinking, but from a formal point of view, this move is only "similar", while It's not like that step 'five road sharp rush' gods are all just.

Of course, Li Xiangping thinks now that Mr. Wu is like AlphaGo, or that AlphaGo learns from Mr. Wu, these are meaningless. When Go AI starts to learn from human chess manuals, it has already left humans behind. very far.

It is also for this reason that when Ke Jie Ke Shaoxia watched Li Shishi VS AlphaGo, he threw the sentence "The dog can beat Li Shishi, but it can't beat me".

It was only at the time. Li Xiangping did not think that Ke Shaoxia's words were arrogant. After all, a generation of dogs still bears a deep imprint of human Go. If the dog hadn't evolved, Li Xiangping thought it was not impossible for humans to win it.

Back to today's game, when Li Xiangping made this "first dog move", Li Canghao began to slow down his speed. By the middle of 11:30 in the morning, there were more than 30 moves overall. During this one and a half hour competition, Li Canghao took half as much time as Li Xiangping.

Of course, the chess players who watch the game in the spectator room don't care about the time of the game, they only watch the progress of the game.

At this time, whether it was a professional chess player watching the game or an ordinary melon eater watching the game online, everyone expressed their emotions:

"Tsk tsk, it seems that Shifo will also be angry, and it is understandable to think about it. In the face of Li Xiangping's bullying tactics, he is a common hand and a first-timer, so it may not work if he is not held accountable. , Now, Li Canghao should think about attacking a large piece of black chess."

In the game room, Li Xiangping, who was fully involved in the chess game, of course also guessed what his opponent was thinking. He knew that Da Li must be thinking of an attack route now.

Li Xiangping laughed silently!

Although there are only more than 30 moves at present, Li Xiangping knows in his heart that the game strategy he and Lao Shi jointly designed today should be considered a success.

Leaving aside the fact that the attack itself is not what Li Canghao said he is good at. For this chess shape alone, it is not so good at attacking.

"Alas, the inertia and misunderstanding of human thinking"

Li Xiangping pretended to be so sighed.

At this time, Li Xiangping even saw victory beckoning to him.

So what can this game of chess be like? The chess game continues step by step.

Come up with a chapter before the New Year's Eve dinner. I wish all book friends a happy New Year's Eve and family reunion!

In addition: In principle, it will be updated continuously during the Chinese New Year, but everyone knows that during the Chinese New Year, whether it can be updated stably, Uncle Seven cannot guarantee it now. Please understand this, and at the same time in the new year, please continue to support Uncle Seven, thank you!

(end of this chapter)

Chapter end

Report