2019年12月17日星期二

Can robots and humans live in harmony?

In early 2017, the upgraded Master of AlphaGO of Google artificial intelligence defeated all the top players in the world with a score of 60:0, causing global concern. If we say that in March 2016 Alpha Dog defeated South Korea's first player Lee Se-shi with a 4:1 victory, it was more of an event in the chess world. The significance of this "slaughter" is obviously not limited to this. Why is AI so strong, and in what areas will they challenge humanity, and will our future be full of a sense of crisis? Master's “Extreme Place” At the end of 2016, the psychology of the chess world on Alpha dogs is dissipating. This is not because more AI players have appeared on the online Go website – such as DEEPZEN in Japan and China’s "Extraordinary art" - and more importantly, human players have a chance to win each other when they compete against each other. In particular, when China's top player Ke Jie once won the upgrade of the “Extraordinary Arts” version of “Crime Day” (which is considered to be equal to the Alpha Dog) and declared that an AI loophole can be found, the human chess player rejoices. However, the Master is on this background and harvests without any suspense. After the game, Nie Weiping, a chess player, said, "We can't make mistakes like computers do," and Ke Jie sighs again and again. "Mankind's actual practice has evolved for thousands of years. The computer tells us that everything is wrong..." Where is the strong? Dr. Huang Shijie of the Alpha Dog Development Team stated that its “excellence” lies in the convolutional neural network backed by multiple sets of supercomputers. This allows the Master to not only have a “strategic network” similar to the human brain (selecting a player’s hand The best method is to have a "value network" that is difficult for humans to judge (determine the precise winning rate for each step of a static game). At the same time, Master can also increase his power by playing hundreds of thousands of games each day - the maximum number of moves a human player can perform in a year. As a result, Master was able to put aside the exhaustive analysis that was twice as effective, and responded with lightning-fast efficiency to make human players difficult to parry. Computers will “learn” cultural ideas For computers, all tasks are logical operations, and algorithms and calculations are the main indicators for improving efficiency. The optimization algorithm can solve the problem more easily, and increasing the calculation power can achieve more results in a unit of time. When an activity has a logical operating characteristic, it can be submitted to the computer for execution; the more rules of change are mastered, the easier it is for the computer to be invincible. Chess is an example. Since chess moves are relatively simple, today's game programs installed on mobile phones already have the power of international special masters. Professional-level chess programs (such as Rybka) are almost all the way to go, and they have the upper hand when confronting human players. The intuition of ordinary people is that it is normal for the computer to lose in terms of computing, and that humanity in terms of cultural creativity is always the winner. In fact, literary and artistic creation activities can also be converted into logical operations through algorithms. The famous science fiction writer Liu Cixin once developed a poetic writing software and set up to create a three-line poem. The rhyme foot is a. Press “Generate” to jump out the result: “Vodka, please atomize it! Cliff, please magnetize it! We are all bargaining chips and we have to struggle!” And if a “holy grail” is established in this area, it is undoubtedly an attempt to create Borges’s “Babelian Library”. This novelist’s imaginary library contains all the texts combined with different characters, numbers, and punctuation marks—both the books we have read and the books that have disappeared in the dust of history. Books that haven't been written yet... Computers create this library by listing all combinations. Continuously optimized algorithms will reduce stress by eliminating a large number of meaningless results.
Contemporary computers have not been able to build this library, just as they have not been able to exhaust the variation of the Go routine. But the performance of the processor has developed at an exponential rate. In 1946, the human computer's first computer "Anyake" was even less powerful than a hand-held calculator. How can we imagine that a century-old computer will be powerful? To what extent? What is certain is that once the computer can be “created”, it will inevitably overcome the humanities and arts fields such as music and painting. The exclusive domain of human creative ideas will encounter the invasion of machine logic. The future AI (Artificial Intelligence) symbiotic with AI is strong and weak. Weak artificial intelligence is only good at a specific area, such as driving, assembling or vacuuming. Alpha dog also belongs to this, but its essence is a deep learning engine, not just playing chess. According to Google's blueprint, it will be used to assist in the judgment of early illness and other areas. Strong artificial intelligence will have human-like language, emotion, and thinking ability. They will surpass humanity in many aspects. Therefore, the significance of R&D is not to solve a specific problem, but to use them to create more horizons. The future society must be a symbiotic relationship between humans and AI. However, on the relationship between the two, three different perspectives can be combed. The first view is represented by Jiang Qiping, editor-in-chief of the Internet Weekly, and believes that robots will only develop in the direction of weak artificial intelligence, and that humans will fully get rid of physical labor with their help. The second view is represented by the American Internet thinker Kevin Kelly. He affirmed that strong artificial intelligence will emerge and that they will live in harmony with humans. In The Book of Necessity, he mentioned: “The things we create make us better people. At the same time, we cannot do our own invention.” The third viewpoint is represented by British scientist Stephen Hawking. It is considered that it is more than worth worrying about whether artificial intelligence can be completely controlled than the question of who controls artificial intelligence. American scholar Ray Kurzweil goes further. He predicted in the book “Beginning to Be Closer” that artificial intelligence will surpass human wisdom and subvert the entire civilization in 2045: “The intelligence that will emerge in the future will continue to represent human civilization... ... The computers of the future are humans—even if they are abiotic.” The fog of civilization made it difficult for us to look to the future. However, most people certainly hope that Alpha dogs don't learn to stand up too early because we want to see the live World Cup. Artificial intelligence, English abbreviation is AI. It is a new technical science that studies and develops the theory, methods, technologies, and application systems for simulating, extending, and expanding human intelligence. Research in this area includes robots, language recognition, image recognition, natural language processing, and expert systems. The industry stated that the technological products brought by artificial intelligence in the future will be the “container” of human wisdom.

Pad Mounted Transformer

Pad Mounted Transformer,Single Phase Transformer,Oil Immersed Power Transformer

Bao Ding Tian Wei Group Transformer Co., Ltd. , http://www.jspowertransformer.com

没有评论:

发表评论