A Go Legend

I was very happy to see that the author of Go Dream has published a new book in Longkong. The latest chapter is in one breath. I feel that the language is much better than Go Dream. . But with the current setting of the spirit behind the chess spirit, it might be impossible to kill a dog. . Because to say the least, even if Shi's skills after adapting to modern rules reach the peak and can defeat the strongest dog at present, the speed of artificial intelligence becoming stronger is beyond the reach of humans. The news will probably report like this: Ten years ago, the chess master easily defeated the dog Zero. Three years later, the latest dog's winning rate against the chess master was more than 90%. The latest version of the dog has a 100% winning rate against any human. . If this were to be regarded as a novel, it would undoubtedly be hopeless. So either treat dogs as tools instead of opponents, such as running, no one competes with cars for speed, such as lifting weights, no one competes with excavators for lifting weights, there is no need. . Why does Go have to be played against dogs Or you can only use dogs to control dogs, further strengthen the system, and obtain a soul Go computer. This option does not match the current writing style, so just take a look. . The following is my comment on Longkong: "Because the rules of ancient chess are different... It's hard to say whether it's strong or weak... It's hard to falsify it if you brag about ancient chess But killing dogs is still nonsense... Dogs are obviously not on the same level as humans I think it's set up like this, get a chess saint system, and then get the soul of the chess player by doing tasks. According to the tasks Different levels and numbers of chess player souls are obtained, and then this system can string the obtained chess players into soul versions of super computers. This makes sense. As the number and quality of chess player souls increase, one's own soul computer will become stronger step by step.

Nowadays, human beings rely on imitating dog tricks to win gold and silver hats. They can achieve almost no loss of winning rate in the first forty or fifty moves. However, after the opening sequence was completed, the game collapsed again. All kinds of unbelievable tricks in the middle game instantly destroyed the human chess players. When it comes to whether a dog's judgment can be learned, it can be said that it can and cannot be learned. It can be because dogs can entangle people with several options. Each of them produces thirty or forty reference pictures. People can understand why the dog chooses the way they do after comparing them. However, it cannot be because it only takes a few seconds for a dog to produce these pictures, and it may not be possible for a person to complete the calculation in dozens of minutes, and it is very likely to be missed. It is even more difficult to compare two or even more pictures in the brain that only exist in imagination. After all, the performance of the human brain is far different from that of computers. After all, the foundation of chess power is calculation power. However, whether it is eye-catching or duel killing, at present, dogs can completely crush humans. In duel-killing, humans may still have a little chance. Dogs still have some shortcomings in handling super large blocks of life and death. Humans should not even think about Kung Fu chess. In fact, dogs are much more sensitive to the eye than humans. They can calculate the outcome of victory or defeat with more than a hundred moves. Just because it is as stable as a dog mode, humans have very few opportunities to learn from it. A lot of scattered pieces are used to create soldiers, or depth bombs are inserted, and both sides are inserted at the same time. This is common in dog charts, but at first glance, it is not a carbon-based thinking, because this kind of actual life and death, which is more difficult than the devil's hand, is almost impossible for people to discover in the actual game, let alone play it out.

On January 3, 2017, master (alphago v18) played against maker (Park Yeon-hwan) on Yefox. This chess record can be found on Yehu, and there are also explanations by professional chess players. The master held the white chess piece and defeated Park Yeon-hwan's Chinese style. Park Yeon-hwan did not choose to connect in the game, and in the end he was defeated by the routine of flying behind and pressing back. Bai Qi did his job easily, proving that the pressure of Dao Ce recognized by the AI ​​was his own, and that counterattack was unreasonable. Of course, AI does not think that there are many good players in some famous ancient chess games. Jueyi thinks that the good players with red ears are incomprehensible. AI believes that chess with unclear intentions (too many human thoughts) is not a good move. The most powerful thing about AI is that it can always make the first move. There are also flaws. The exquisite handicap chess game proves that when the AI ​​lags behind, it is likely to fall into human traps and routines. If the version is not upgraded, it will be the second time.

Let's talk about talent first. Although Shi Xiangxia is extremely talented, and Gu Li is also a genius who emerges every ten years, Gu Li must be worse than Shi Xiangxia? Well, even if Shi Xiangxia's talent is higher than Gu Li's. There are only four ways for chess players to improve their chess skills: teaching from masters, studying chess records, competing with opponents, and studying on their own. Whether it is the quality of teachers, the number of chess records, or the strength of competitors, Shi Xiangxia has far less resources than Gu Li. To put it bluntly, the number of serious games that Shi Xiangxia has played in his life may not be as many as those played by Gu Li in a year. Not to mention, Shi Xiangxia had never heard of or seen many of the chess records that Gu Li had studied. Under such circumstances, I think it is a bit unreasonable for Shi Xiangxia to beat Guli violently in fast chess. Shi Xiangxia's situation is different from Zuo Wei's. Zuo Wei's soul has existed for thousands of years, and he has been playing chess with himself to improve his chess skills. Later, he possessed Torajiro, and he was not too out of touch with modern Go. Therefore, his superiority over all living beings has a practical basis. Why does Shi Xiangxia have such high chess skills? I don't believe that a person who plays ten games a year can compete with a national player who plays ten games a day. No matter how talented he is, it is absolutely impossible. For Go, creative thinking is important, but memory and proficiency are the foundation. In these aspects, ancient players and modern players are not at the same level. If you write like this, you are disrespecting Go and the players who have worked tirelessly to improve their Ichimoku skills. Go is certainly an art of genius, but without enough effort, it is impossible to reach the top.

The writing is very good. Don't listen to those people's nonsense. There are a few people who read the novel who really understand Go. How can an amateur chess player without a golden finger become a legendary chess player? As long as the novel is good-looking, don't open a harem randomly!

Complaints from all book readers I hope the author will stop talking about other things. Are there really people who read this book to see the protagonist living in the city as a time traveler? First, I wrote about real estate. The author is engaged in real estate, so I couldn't bear it. Later, I wrote about TV dramas. Although it felt awkward, it was still related to Go, so I jumped into watching it. Then I wrote about the entertainment industry. I wanted to follow it for so long, and it would be a pity to give up. Now I am writing about financial investment, but I can't bear it anymore, so I say goodbye! If we want to read this kind of urban pretense novel about time travel, why not find an urban novel? The author doesn't care what we want to see! Obviously the current update speed is already very slow, but if I still have to write about things other than the main story, I will probably be hopeless in dogfighting within a year. I took a look at the recent updates and found that most of them are about things that have nothing to do with Go. Although the plot does need some relaxation, it's almost like the other way around now, with the plot focusing on urban life and the small heads on the chessboard. Actually, I think it's better to read without any updates than to read the author perfunctorily writing something he doesn't want to do, but the author doesn't care

Except for the part about playing Go, the description of all other non-Go content, especially the description of the protagonist's mentality and inner thoughts, is very jarring. Schizophrenia + childishness are not enough to describe this sense of incongruity.

One thing to say, I plan to continue reading this book, but there are some issues that I really don't like. I haven't seen the later part yet, but in the first 30 chapters, I have now seen the LG Cup qualifiers before the match against Xiao Lin. The protagonist is a salted fish without dreams. It sounds good to say salted fish, but it sounds bad to say that he is a gangster. With Shi Xiangxia's personality charm and revenge, The pursuit is in sharp contrast. First of all, the scumbag character before the time travel was very unpleasant. Not to mention that after the time travel, he had no ambitions and just wanted to eat his old age. He also walked the dog in the game (referring to using AI to play for him, using Shi Xiangxia to cheat and feel proud, and not concentrating on studying Go). I know that there are such people in reality, but this is a novel. This kind of protagonist is very disturbing to watch, and there must be a change in the later period, otherwise the book cannot be written. So why not make the protagonist more likable from the beginning? Even if it is not set as the rich second generation, a child from a family of ordinary Go enthusiasts will have no impact on this book. It's a little depressing. Please forgive me if it offends me.

Many people don't understand Go AI at all, so here's a brief introduction: First of all, the area that humans are responsible for calculations is limited; this aspect has long surpassed the computing power of Superintelligence. This is the fundamental reason why we lost chess. But again, if you don't give AI enough computing power, magic programs will be useless. Secondly, why has Go AI defeated humans only in the past few years when its computing power has already surpassed that of Go? Mainly it's an algorithm problem. Only computing power and appropriate algorithms can produce effective output, and exhaustive methods will not work. The learning algorithm of artificial intelligence has given an idea, so a lot of AI that surpasses humans have emerged. Many people here should understand it at a glance; yes, AI learns Go using the same learning methods as humans. It does not develop from "0", but through "feeding scores" and other methods. Like every chess player, it is the culmination of the past two thousand years of Go. (This is a bit exaggerated. After all, the current Go rules are actually only newly formed in the past hundred years, but they are indeed inherited from the past.) Note: Different rules will lead to different favorable setting stones and algorithms Therefore, Go AI surpasses humans in both computing power and learning speed. Even if there is a "wonderful player" that surpasses it, it will quickly iterate past him. However, they still have two major problems: First, they will "get old"; although they can create new stone-setting techniques by playing against each other, they will also become "saturated" after a certain level. Because that is already the current "optimal solution", in other words, it will "stop" like a human. This gives future generations the possibility to surpass it. (But if people want to surpass it, they must first close the distance; to close the distance with it, of course, is to learn it and use it to improve one's chess skills, just like it was at the beginning of its birth; don't you just cross the river by feeling the eagle sauce) Second, the first problem cannot be solved through iteration. This makes it impossible to become Him, the God of Go. The first problem will be faced by any "mature" Go AI; to solve it, it can only be solved by modifying the underlying weights and rebuilding one, but it may not be successful. Simply put, the next one is not necessarily better than the last one. This is embarrassing because each one costs a lot. In other words, in the future, it is more likely for humans to make a breakthrough "god move" than for Go AI (the cost is lower, because every mature chess player has the opportunity to do it, which is exactly the opposite of AI). (However, it cannot be denied that most people cannot even approach the general level of AI; it's just that there are only a handful of top chess players themselves) Finally, Go AI actually has two major benefits for human chess players. One is that it can quickly reach the current upper limit, and the other is that it can more quickly verify whether some new setting stones are feasible.

When writing about Go, just keep it simple. You don't have to show off all kinds of connections in the capital and look like a Beijing oil kid.