Chapter 219: Exercising Off-site Advisory Rights (7500 words long!)
"Google's AI defeats the world Go champion!"
"AlphaGo defeated Lee Se-sil in the middle game, and Lee Se-sil resigned after 186 moves!"
"Lee Se-sil: I underestimated my opponent in the first game. AlphaGo is stronger than I thought, but it's not invincible."
"AlphaGo's initial victory may herald the advent of the era of artificial intelligence!"
"It's been revealed that Gu Gou once considered challenging Gu Jie. Insider information confirms that Lee Se-sil will definitely lose!"
“AlphaGo defeats Lee Se-sil: The significance of the robot’s victory needs no exaggeration.”
"Intelligence vs. Artificial Intelligence: Are Humans Both Losers and Winners?"
"AlphaGo has demonstrated a professional level of play, but it's not yet at the level of crushing humans."
"Gu Jie: Artificial intelligence will eventually completely defeat humans, but Lee Se-sil should be able to win later."
"Don't be pessimistic. Rumor has it that Gugou and Lee Se-sil had a pre-match agreement that they couldn't use ko. AI can't crack ko, and it can't handle imitation chess."
"Faith Collapse: What Industry Will Be the Next to Be Ended by AI?"
“Will artificial intelligence destroy humanity?”
What is deep learning? Analyzing the history and development of alphaGo.
A single stone can cause a thousand ripples. As soon as the news of Google's alphaGo defeating Lee Se-sil in the first game came out, it quickly occupied the front pages of almost all media.
Even the Anglo-American media, which initially reported little on this match, reported the specific process of the battle in a rare and large-scale and detailed manner.
After all, no one expected that artificial intelligence had quietly conquered the field of Go, a field generally believed to be one where humans could never be defeated by machines!
Even last year, the strongest Go program Zen was still defeated by some top players even if it was given a 9-stone handicap.
At that time, Zen was able to beat some amateur players, which was already an amazing improvement.
And in just one year, the previously unheard-of AlphaGo actually defeated the strongest among humans!?
There are those who question it, those who advocate it, and those who panic, but there is no doubt that alphaGo is popular, and AI is popular!
"Wow! The Woking system is absolutely amazing! All 56 selected stocks have risen to the expected prices, with an average return of 11.8%! Thank you AlphaGo, thank you Li Shishi!" Lu Xuechang excitedly waved his fist.
The day before yesterday, the Woking system gave a recommended purchase portfolio strategy, almost all of which were technology stocks.
Luo Ziming was also filled with shock. "It's really amazing! I was thinking of picking Google as a backup, but it turned out not to have a single stock on the Nasdaq 100 Index! Isn't that too bold?"
"Yeah, who would have thought that Google only rose 1.65%? And Nvidia even fell?"
"Because Google disclosed that it did not use Nvidia GPUs, but instead used its own TPU."
"Google is too big. Its total market capitalization is already $530 billion, and the price of a single share is over $730. The cost is too high."
"So it's these newly listed AI companies that are seeing the most rapid growth. Look at soundAI, it's up 19% today alone."
"It's too bold. This combination strategy is still too bold. The Woking system also judges the risk of this transaction as high risk. It still needs to be adjusted back later. I decided to adjust the position and structure and add Google and Nvidia." Huang Xiang shook his head.
"Lao Fang just called this morning and said not to buy Google yet, but Nvidia is fine."
"Fuck, he's nothing. I have the money, I can buy whoever I want!"
"He's the money man."
In just less than twenty days, the No. 4041 Fund has grown to 16 million Malaysian dollars, and after today, the amount in the account will be close to 18 million Malaysian dollars, an increase of 6 million Malaysian dollars compared to the end of last month!
50% profit in a single month!
If the interest is compounded over a year, wouldn't this 12 million dollars become 1 billion dollars?
Can each of us get tens of millions of dollars?
Fang Yu thought to himself, "What are you dreaming of? Are you really treating the SEC (Securities and Exchange Commission) and the DOJ (Department of Financial Regulation) like a mere decoration? If you really dare to do that, the Federal Reserve and the FBI might even intervene in the investigation."
Moreover, even if he was not afraid of being investigated, it was impossible for the Woking system to really increase the principal by more than a hundred times in a year.
Fang Yu has conducted simulation experiments. Due to the current computing power limitations, the data sampling of the simulated world is very rough. Coupled with the problem of data pollution, the virtual divination of the Woking system is unable to find out the stocks or futures with the highest fluctuations in the next few days.
The Woking system relies on virtual divination to define a large range, and then within this large range, it selects suitable investment stocks based on the internal algorithm of the Orange model.
Moreover, the target price given by the Woking system is not that accurate, and on-site traders still need to make judgments based on actual market feedback.
Otherwise, in the case of short-term trading, once the target price of the Woking system fails, huge losses may occur, which cannot be solved in a short period of time.
In addition, as the controlled assets increase, the marginal profit will definitely decline. This is the law of the market and there is no way to change it.
Because the total market size is so large, when the fund size reaches a certain level, every move will cause market fluctuations and returns are bound to decline.
However, Fang Yu's more than 10 million dollars cannot be considered a large scale. To be precise, there is no question of achieving marginal returns before the fund size reaches 10 billion dollars.
Although a hundredfold return is impossible, a dozen times should be about the same.
However, Fang Yu's attention was not on the stock market, because the second battle between Li Shishi and AlphaGo was about to begin!
——
If no one recognized Hassabis on the first day, then on the second day, when he walked from the seventh floor to the spiral staircase on the sixth floor, he had to cover his face and dodge and rush a few steps before he could get rid of the reporters' entanglement and return to the temporary control room on the sixth floor.
Unlike the first game, in order to avoid being disturbed by media reporters, the DeepMind team set up a temporary control room on the sixth floor, where all members monitored the background operation of AlphaGo and watched the live broadcast monitoring on the scene.
They just visited the capital of Goryeo this morning. The biggest topic in Goryeo right now is the man-machine battle between Lee Se-sil and AlphaGo.
When they saw the big screens on the streets, newspaper headlines, and even the bus stops where you could watch videos showing this battle of the century, everyone on the DeepMind team could see their heartfelt joy on their faces!
Yes, we are making history!
We are the pioneers of the artificial intelligence era!
Compared to the ease at the beginning of the first battle, Li Shishi's expression was obviously much more serious in the second battle, and his moves were more cautious.
"Oh, Li Shishi is being too conservative." Gu Jie couldn't help but slap his thigh when he saw Li Shishi's tenth move for white. "Why not play three-three?"
"Maybe he feels that he doesn't want to use your moves?" Chen Ying joked, and the other two male commentators nodded frequently.
"I said, that was something I learned from Orange. It's not my move." Gu Jie quickly shook his head to deny it.
The male commentator next to him cheered: "But you are indeed the first one to use this routine in a formal game, right?"
Gu Jie changed the subject: "The Great Zhou Chess Academy has already discussed this move extensively. If White plays the new point-three-three opening, it can significantly increase the chances of winning. Followed by a two-space jump, the situation opens up immediately. It all depends on how the AI pursues the opponent."
"But now Li Shishi still uses the Xing Xiaomu opening. Now the AI is obviously using the big circle opening. If the AI makes some targeted changes later, Li Shishi will be very passive." Gu Jie placed the chess pieces on the chessboard.
"Huh? Breaking the right side? The pattern is not completed?" Gu Jie had just finished setting up the chess pieces when he saw Huang Shijie making AlphaGo's 13th move.
The 13th move is a move that breaks away from the fixed pattern. In essence, Black loses the initiative. At this time, White can completely punish Black by breaking the edge where Black just made the fixed pattern, making Black passive locally.
If an ordinary person made this move, he would definitely be considered an amateur player who doesn't understand chess. But this is AlphaGo, which just defeated Lee Se-sil yesterday. How could such a mistake be made?
Obviously, Li Shishi was also very cautious about the unexpected changes in the black chess pieces.
After a long consideration, Li Shishi placed the white piece on the star position in the middle below with a snap.
"Ah, this is to cut off the AI's retreat." Gu Jie praised and moved the chess on the board. This move was equivalent to Li Shishi predicting the AI's prediction and blocking the AI's response route.
Less than a minute later, AlphaGo made its 15th move on the screen: Stab.
"Stab?" Almost all the professional chess players in the studio were surprised.
It's not that this move was good, but that it was a completely vulgar one.
The so-called common moves are those often used by beginners or chess players with lower skills. These moves often seem effective, but they are very short-sighted and ignore the overall layout and long-term impact. To put it bluntly, they are picking up sesame seeds and losing watermelons.
Even if a young chess player comes from a formal training institution, if he makes this move, he will be slapped on the palm by the teacher!
After this move, because the black chess piece has the first move, it is equivalent to exercising the right of first move, and the white chess piece is indeed unable to resist. However, the black chess piece in the upper right corner also loses many potential changes and is finalized too early, which is very unfavorable for future layout.
What does AlphaGo mean?
AI has just been born, and human chess players have not yet truly understood AI's thinking. Quite a number of commentators believe that AlphaGo made a wrong move.
Only Gu Jie looked thoughtful.
"I think I can understand AlphaGo's intention." Gu Jie pushed his glasses and squeezed his eyes unconsciously.
"Oh?" Chen Ying was curious. "What's your intention?"
"Li Shishi's two responses just now were actually intended to undermine the AI's intentions, but he overlooked one point: AI has no intentions; it only coldly calculates gains and losses, wins and losses.
"The same is true of this move, 15 stabs. We consider it a vulgar move because it was prematurely finalized. But the question is, is premature finalization wrong? What is our standard for right and wrong?"
"The black piece in the upper right corner has been fixed, gaining territory. Isn't that a good thing? As long as the AI determines that this exchange is worthwhile, it has no subjective biases about vulgarity or elegance; it only calculates whether it is appropriate or not."
Gu Jie calmly placed the chess pieces. No one, including himself, knew that, up to now, he was the top chess player who had played against AI the most times. Unconsciously, he began to understand many of the AI's ways of thinking and subsequently moved closer to AI.
The next dozen or so moves were all set patterns, until the 29th move, when AlphaGo's black chess pieces once again surprised the audience a little.
In the words of Liu Weiping, a veteran of Big Surround Go: After seeing this move, he couldn't help but take off his hat and pay tribute to AlphaGo.
Although this move deviates from the existing response plan, it is a move that humans can fully understand. It also fully demonstrates AlphaGo's dominance. Even if a human makes this move, he will be praised for his courage.
According to the understanding of human chess players, the black chess piece should be placed on 3J, but AlphaGo was not conservative at all and directly placed it on the star position in the upper right corner of 3J.
It was just a small move, but it meant that AlphaGo not only had its eyes on the territory on the right side, but also showed its ambition for the middle part without reservation!
You know, in this case, the amount of calculation for subsequent chess moves is more than ten times greater than the previous mature response plan!
Li Shishi's response on the 30th move was standard. So far, almost all chess players believe that the winning rate of both sides is still 50-50.
"53% win rate." David Silva and Hassabis crossed their arms and pointed at several monitors hanging on the wall. The complex charts on the screen showed that after Lee Sesil made this standard move, AlphaGo's win rate increased by 2 percentage points.
It was not until Li Shishi's 36th move that Li Shishi nodded slightly, seeming to be quite satisfied with his performance in this game.
I really want to smoke a cigarette.
Li Shishi touched his pocket, stood up, walked out of the game room, and went to the outdoor platform on the sixth floor to smoke.
Just as Li Shishi lit a cigarette, AlphaGo gave the landing point of the 37th move.
Five shoulder charge!
Since the previous few moves were relatively ordinary, the professional chess players and hosts in each studio were chatting. Gu Jie's studio was no exception. The four hosts were still chatting about the 15-stab vulgar move that had just happened.
"It's not surprising that AlphaGo was able to make this move. High-level human games, from ancient times to the present, total only a few hundred thousand, far too few for machines to statistically analyze. However, low-level games are played online in perhaps hundreds of thousands of times a day. This tactic is not uncommon in low-level games, so it's not surprising that a computer could learn it."
Professor Liu Ziqing from the School of Computer Science at Xinyou University also served as the host and commentator, providing professional advice for the commentary of this game.
"What fucking nonsense!" Hua Zecheng snorted as he watched the live commentary. "You're talking about statistical analysis? You haven't even figured out the basic principles of AlphaGo. I don't know how you became a professor."
Fang Yu chuckled and kept sending WeChat messages. "Everyone has their own expertise. Professor Liu specializes in tree search, so it's normal that he hasn't been exposed to deep learning."
Gu Jie, who was on the livestream, was the first to spot AlphaGo's move. He gestured to the female host and brought the conversation back to the game: "This move is incredible! A five-way shoulder charge. A human would definitely not make this move. When I was playing against Orange..."
As Gu Jie was speaking, he began to think deeply.
"Gu Jie?" Chen Ying pushed Gu Jie calmly. This is a live broadcast. Why are you in a daze? No matter whether what you say is right or wrong, you have to speak. Huanshi paid you so much, not to let you learn chess.
Gu Jie came back to his senses, smiled embarrassedly, continued to finish the chess game, and said a few words to get by, but it was obvious that everyone could see that Gu Jie was absent-minded.
After AlphaGo made its move, someone from the Goryeo Chess Academy went to the sixth-floor platform to inform Li Shishi. Li Shishi hurried back, and before he even sat down, he saw AlphaGo's five-way shoulder charge!
Li Shishi took a deep breath, an incredible look appeared in his eyes. He smiled first, and then fell into deep thought.
In fact, this move is not without a solution. On the screen of the DeepMind team, AlphaGo's winning rate has not changed at all.
AlphaGo's winning rate is calculated based on the results of self-play, and whether humans can make the moves recommended by AlphaGo depends entirely on the player's chess skills.
"Orange? This move is too similar to an orange." Li Shishi was also thinking in his heart.
"Could it be that Orange is really also an artificial intelligence?"
"If that's the case, even though I lost the game of chess with Orange in the middle, it wasn't that frustrating."
"It's really frustrating."
Li Shishi suddenly realized that he had completely misunderstood AlphaGo.
He was too concerned that his opponent was a computer program and did not regard him as a real chess player.
Previously, when he played against Orange, due to the difference in chess skills, although he lost in the middle game, he still had the advantage for a while. For him, the situation on the field was very enjoyable, at least in the first sixty moves.
Given the huge gap in chess skills, he could accept his loss with a clear conscience.
But now, he was only focused on destroying the computer's intentions, focusing on his opponent rather than his own chess pieces.
This kind of chess, even if you win, is a shame!
I am Lee Se-sil!
I am the best in the world!
I represent the dignity of chess players!
Li Shishi's eyes suddenly became sharp.
In that case, let’s play a round!
“Bang!”
Li Shishi picked up a white chess piece and placed it on top of the white chess piece without hesitation.
Seeing Li Shishi's response, Gu Jie's eyes widened and his body straightened up suddenly.
Is Lao Li going to fight?
Isn’t it a bit early?
AlphaGo jumped without hesitation.
Li Shishi, on the other hand, seemed to have predicted this move in advance. He ignored AlphaGo's small jump and put the chess piece on the left side of the board without hesitation!
After this move, AlphaGo's winning rate dropped sharply to 50% on the screen in front of Hassabis!
What’s going on!?
"This move..." Gu Jie took a breath, "This is exactly Yuzi's move. It seems that Lao Li is really going to attack head-on."
Obviously, this move was also beyond AlphaGo's predictions. AlphaGo rarely ran for more than two minutes before giving the next step's placement.
AlphaGo also chose to fight head-on!
"Snap"
"Snap"
"Snap"
"Snap"
As the sound of chess pieces dropping continuously came, all the spectators holding their breath unconsciously.
"Brilliant! So brilliant!" Liu Weiping's nose flushed with excitement. He couldn't help but cheer as he watched Lee Se-sil's 78th move on the board. "This is Go! This series of offensive and defensive moves by both sides just now is arguably the best in the last ten years, no, the last twenty! Aside from Wu Qingyuan's ten-game match, I can't think of a more exciting match than just now!"
"Li Shishi is worthy of being the champion for many years. Today's game of chess is the best he has ever played in his life!"
Gu Jie was also dumbfounded. Is this still the old Li who he could easily defeat?
By the 134th move, a layer of fine sweat beads had seeped out of Li Shishi's forehead. His face showed pain and his brows were furrowed. He picked up the coffee in front of him and took a sip, then took a tissue to wipe the sweat off his face.
The current level of chess play has completely exceeded everyone's expectations, including Lee Se-sil.
It’s so powerful, artificial intelligence is so powerful!
Although it seems that the two sides are still in a stalemate, Li Shishi is very clear that if he is even slightly careless, the dog on the opposite side will tear him to pieces!
In the previous game, Li Shishi admitted that he underestimated his opponent at first, and his mentality became unbalanced when he fell behind, which eventually led to his failure.
But he didn't think AlphaGo's strength had really surpassed his, he just felt that he didn't perform well.
The same is true for the first thirty-seven moves of this game.
But after adjusting his condition in the thirty-eight moves, he can proudly declare that the forty-odd moves he has made so far are the best he has ever made!
Even if Gu Jie, who has always had an absolute advantage over him in terms of record, came, he would still lose to him!
Especially his 112th move was a move that would make him proud for the rest of his life.
However, no matter how fierce his attack was, or how dead but not stiff his zombie flow was, he could not gain any real advantage from AlphaGo.
So powerful!
Become stronger when facing a strong opponent!
When I was not in good shape yesterday, this AI did not seem that powerful. But in today's game, I played my best and made the moves I was most satisfied with, but I still couldn't win?
No, it's only the late middle game now, not the endgame yet, I will definitely have a chance.
Li Shishi's eyes flashed, and with a snap, the chess piece fell on the empty space on the chessboard.
Robbery!
"Robbery! There must be something wrong with how AI handles robbery! Isn't that right, Gu Jie?" Chen Ying also asked Gu Jie excitedly.
Gu Jie hesitated. "Perhaps, but the previous chess pattern required considerable calculations, as AlphaGo's performance showed. It's unclear whether ko can trigger AlphaGo's bug. Let's see how AlphaGo responds. Oh, it takes it, and Li Shishi follows suit. AlphaGo cuts off the capture. Ah, AlphaGo's response to the ko is excellent. Ah, it pounces. This pounce is excellent, and no ko material is wasted at all."
The robbery that Li Shishi painstakingly provoked was easily resolved by AlphaGo.
"It's official game, let's see who wins in the end. The situation is quite complicated, and it seems that the gap between the two is not that big at the moment." Gu Jie took off his glasses and rubbed his eyes.
Li Shishi, who had already entered the endgame, had an abnormally flushed face.
He could feel that the high speed of his brain just now seemed to have overdrawn his energy and spirit.
Even after playing this game for so long, he still had no confidence in defeating AlphaGo.
"Win by half a point." Hassabis looked at the AlphaGo screen calmly and already knew the result in advance.
In fact, at the 161st move of the whole game, he knew that AlphaGo had already won.
Because at that time AlphaGo's winning rate had exceeded 75%.
It’s just that the subsequent calculations are too complicated and far beyond human capabilities, so human chess players cannot see them.
"Li Shishi lost." As the space on the board became smaller and smaller, Gu Jie had already calculated the final result in advance, but he did not look too disappointed.
"AlphaGo is too strong. Even I might not be able to beat him." Gu Jie smiled awkwardly. "You're slapping me in the face again. Yesterday I said AlphaGo couldn't beat me."
"Even if we defeat Li Shishi, it doesn't mean humanity has failed," Liu Ziqing consoled him. "Didn't that Orange on the internet defeat you before? Do you think he has a chance of winning against AlphaGo?"
Gu Jie smiled and said, "Even if Orange beats AlphaGo, it can't be said that humans have won."
"Oh? You think Orange can win? If they can win, why isn't it a victory for humans?" Chen Ying asked curiously.
Gu Jie hesitated for a moment and said, "I am now basically certain that Orange is also an artificial intelligence."
"The reason I wasn't so upset about Lee Se-sil's defeat today is because this isn't the first time an AI has defeated a top human Go player."
"Orange had already done this more than two months ago."
Is Orange also artificial intelligence?
"Are you sure? Have you received any news?" Chen Ying asked hurriedly.
This is big news! Bigger than a bomb!
Gu Jie said calmly, "No company has contacted me yet. However, considering the two games between AlphaGo and Lee Se-sil, I'm pretty sure Orange is also an AI."
"Their thinking has many similarities, especially in calculations. They don't care about many bold moves that we professional chess players call vulgar or bad, and they all have very good results."
"That's because they're all AI, and they all have strong computing power. Often times, common or bad moves aren't bad, but the calculations after they're made are too complex, beyond human capabilities."
"Artificial intelligence is different. Its computing power is so powerful that it can calculate every step in great detail, turning even the most inept players into skilled ones."
Gu Jie, a young man who just turned eighteen, had no idea how much of a shock his words during the live broadcast would cause!
The chess pieces fell on the chessboard one by one with a clatter. After the final sorting and counting, Li Shishi let out a long sigh.
"I lost."
This time, no one cheered, either at the scene or in the studio. Even Hassabis, who had already come down from the game room on the seventh floor, had a serious look on his face.
He himself did not expect that when he saw AlphaGo defeat Lee Se-sil in his prime, he was not so excited. Instead, he felt a sense of grief as a fellow human being.
"I played this game very well. I may never play this well again, but I lost. I'm sorry."
Under the flash lights, Li Shishi bowed deeply to the reporters in the audience, paused for a few seconds, and stood up again. Everyone could see that his eyes seemed to have turned red.
Li Shishi raised his head and smiled. "In tomorrow's game, I will exercise the off-field suggestion rights stipulated in the agreement. Yes, I will invite foreign players. Please look forward to my performance with them."
Off-site advisory rights?
What's this?
Hassabis and David Silva looked at each other, what was going on?
This item was not originally included in their agreement. It was added temporarily by the Korean Go Association before the start of the competition. If the team loses the first two games, they can get off-field advice, that is, tips, for the next three games.
They think that if Lee Se-sil loses the first two games, it means that AlphaGo's chess skills must be far superior to Lee Se-sil's. Playing chess is not the same as the more people there are, the better it is. Wouldn't it be worse if there are more people talking at the same time?
So they agreed.
They also heard that Lee Se-sil was very dissatisfied when the Goryeo Chess Institute added this rule at the last minute, thinking that the Goryeo Chess Institute did not trust his ability. Later, the Goryeo Chess Institute used some unknown means to make Lee Se-sil compromise.
How could Li Shishi, such a proud person, ask for any off-field advice?
"Boss, how did you get Li Shishi to agree to let Orange play the last three games?" Hua Zecheng was also very puzzled.
In just one week, the Goryeo Chess Institute and Lee Se-sil were dealt with. Isn't this amazing?
"Because I have super powers." Fang Yu took a bite of the apple.
"What superpower?" Hua Zecheng was confused.
"The Goryeo Go Institute is sponsoring 5 million dollars, and Lee Se-sil is personally sponsoring 3 million dollars."
“It’s this money power.”
Fang Yu chewed the apple with a crunching sound.
Damn, why wasn't AlphaGo created by Apple? If it was, how relevant would it be now?
(End of this chapter)
Continue read on readnovelmtl.com
