Reading experience in the intelligent age 1
The content of this book is relatively shallow, basically belonging to the kind of telling stories and browsing. The author's writing, knowledge and experience are excellent, which is not reflected in this book.
The only bright spot is that I didn't know much about the specific relationship between big data and artificial intelligence before. Explain this book briefly.
Traditionally, machine intelligence means that machines can think and solve problems like people. However, because humans have not learned to solve problems in the current scene, they often apply other knowledge to solve this problem, so it is difficult for machines to learn "human" thinking in terms of migration ability. So what is the relationship between big data and machine intelligence? Another way to develop machine intelligence is to completely abandon the road of "thinking like human beings" and directly aim at the ultimate goal of "solving problems". For example, in translation, the previous idea was to let the machine understand all kinds of grammar. Now it is not necessary for the machine to understand grammar, but to "learn" hundreds of thousands of sentences directly, and the dosage changes to achieve qualitative changes. In other words, when you see more and more expressions, the accuracy of translation will continue to improve.
Reading experience in the intelligent age II
This title comes from the preface of this book written by Professor Li Shanyou, the founder of Chaos University, and is borrowed here.
Generally speaking, the main content of this book is to discuss how machine intelligence based on big data develops, how it will affect all aspects of people's lives, and how to deal with it.
Big data has brought profound changes to the world, the most profound is the change of thinking mode-from causal mechanical thinking to relevant thinking. What has changed here is not only the means to solve the problem-now we have the ability to analyze the problem in the complete works, and finally we have the ability to accurately describe the outline of the group on the basis of the complete works, as well as the way to look at the problem-we are no longer "careful assumptions and bold verification", but "reducing assumptions and self-evident data". This transformation actually makes machine intelligence a foundation for becoming a public facility-it is not necessary for everyone to understand how machine intelligence works, but for users, it is only necessary to understand that there is such a tool to find appropriate conclusions in a complete data set. Therefore, this will quickly trigger a commercial revolution.
In fact, this revolution has already begun: in the book, it has been listed that traditional industries, including bar and refrigerator manufacturers, use big data to improve their business conditions. However, there are still more industries exploring how to effectively integrate data and explore new models-first, the problem of data acquisition, and finding an effective way to obtain data, which may not be just nine Niu Yi hairs; Secondly, how to find value from these data still needs a guiding direction; Finally, how to use the rules of data discovery to design a more reasonable model.
Another wonderful content in the book is the last chapter, which discusses the positive and negative effects brought by the intelligent revolution. I think the author actually asked such a question: the intelligent revolution has greatly liberated the productive forces and promoted production, and only fewer people are needed to feed the world. What about the rest? By discussing the historical process of the first, second and third industrial revolutions, the author draws a conclusion that it is not optimistic-only through the role of time can the great liberation of productive forces brought about by the revolution be released. Unfortunately, we are in this historical process, and we are fortunate to be in this historical process.
So the whole book can be said to be a science book. After reading the whole book, you can further understand the most fashionable concepts-big data, machine intelligence, blockchain and so on. , especially big data. But I think it is more like a propaganda book-because it depicts two worlds, one is a world that participates in the revolutionary process and keeps up with the wheel of history, and the other is a world that resists the revolution and is abandoned by history-and the difference between the two cannot be ignored. As for whether it will really be so serious, this is a matter of opinion, but history tells us that at least the first three times were like this. If you have doubts, you'd better pray that there will be something new under the sun.
As for the revolution, there is no doubt about it. Of course, this is known before reading a book. In fact, many people already know that it is difficult not to think about the impact of big data on the world in an environment where the whole society is discussing big data. The utility of this book is only to let readers know how irresistible this process is.
Reading experience in the intelligent age 3
Rereading Dr. Wu Jun's chapter "Intelligent Age-the Turning Point of Technology" and some thoughts. Everyone seems to have such an intuitive perceptual knowledge. Artificial intelligence seems to have entered a rapid development stage from 20 15 and 16. Various conceptual products of artificial intelligence have mushroomed everywhere, especially entrepreneurs, who feel particularly excited and want to create revolutionary products and change the world. At this time, we are actually a little puzzled and confused. Why an artificial intelligence technology that has been developed for more than half a century can be generally accepted and adopted by everyone at this point in time, and everyone has the ability to build an artificial intelligence learning system, showing a breakthrough development trend. We must be wondering what happened behind this, yes!
In the past ten years, another technology has been very popular, and that is big data technology. Someone once described the relationship between big data technology and artificial intelligence technology, saying:
Artificial intelligence is like a baby with unlimited potential, and big data is milk powder for feeding babies. The quantity and quality of milk powder determine the level of infant's intellectual development.
So what are the breakthroughs in big data technology in this decade? Let's look at it from four dimensions, as shown in the following figure:
Reading experience in the intelligent age 4
1. There were four great inventions in ancient China. Why didn't the industrial revolution break out and form the foundation of modern civilized society? The reason goes back to the way of thinking formed by Euclid's Elements of Geometry and Newton's Mathematical Principles of Natural Philosophy: the world is cognizable and can be expressed by simple mathematics with definite laws, and there is a causal relationship. This way of thinking forms a system and permeates natural science and social science, all of which follow these laws.
2. The world in the information age is uncertain and presents uncertain rules. Big data has the characteristics of large quantity, multi-dimension and completeness, which provides people with a brand-new way of thinking, turning intelligent problems into data problems, replacing causal reasoning with strong correlation, and replacing the mechanical thinking mode of bold assumptions and careful verification with the advantages of exhausting examples in big data;
3. The world is uncertain, and all the problems are to solve this uncertainty and turn it into certainty. For example, the certainty of oil exploitation technology solves the modern energy problem, and the certainty of celestial motion law solves the space problem. Nowadays, the uncertainty in the Internet era requires big data+information entropy to solve the problem of certainty.
Reading experience in the intelligent age 5
Reading notes: 320 minutes, Samsung.
Data and information, by filtering useful data to get information.
Analyze the role and development of big data from the perspective of the history of science and technology; From the perspective of intelligent revolution, this paper compares previous revolutions, and analyzes the industry changes and social impacts brought by machine intelligence.
Simple and accurate model and data-driven method. Geocentric theory, Heliocentrism's model. Space shuttles from the Soviet Union and the United States, and optical lenses from Germany and Japan.
The amount of data and the representativeness of samples. Literary abstracts and Gallup's predictions of American presidents Roosevelt and Langdon, and Gallup's predictions of Dewey and Truman.
The basic laws of physics are discovered through statistics.
Probability theory, statistics, the basic knowledge that must be learned. Record your own time, work data, cash flow, etc. They are all accumulating raw data, and then improving themselves and their work through data discovery.
Birds can fly: machines must think like people to get intelligence. SYSTRAN .
Statistics+data. Google translation, speech recognition between Jarinik and IBM, machine learning.
Big data: 3V, rich, diverse and timely. Baidu's "eating goods", Swire's 20 12 election forecast, intelligent traffic management. Big rather than big.
The intelligence problem is a data problem. Deep blue and Kasparov play chess and automatically ask me why and how to play.
It is most important to realize the changes and twists in thinking in the history of artificial intelligence development.
Machine thinking: Euclid, Ptolemy, Newton. The prototype of the mathematical model is obtained through observation, and then the model is refined by using the data. 1 The law of world change is certain. Starting from certainty, we can describe the law clearly with simple formulas or languages. The law should be universally applicable.
Certainty (predictability) and causality
Industrial Revolution, Watt's Steam Engine. Einstein's theory of relativity, penicillin
Mass-energy conversion of three formulas, uncertainty principle of quantum mechanics, entropy
Uncertainty: 1 Too many variables, 2 objective world itself, micro world. information theory
Entropy, information entropy: the measure of information is equal to the amount of uncertainty.
Mutual information: strong correlation. Lawsuits in the United States on smoking is harmful to health, new drug research and development, and search company click patterns.
Shannon's first law and huffman encoding's (Gilder's law) give the most common Chinese characters by short-mouth coding, which makes the average length infinitely close to its information entropy.
Shannon's second law: the propagation speed of information should not exceed the capacity of the channel.
Uncertainty, strong correlation, information theory
Analyze the electricity consumption of smart meters through big data and grow marijuana. Using big data to analyze tax evasion of small enterprises. Target uses big data analysis to recommend promotional information. Amazon big data recommends sales, and there are products that directly push goods. Netflix recommended videos. Google search keyword tips and related searches. Bar shelf renovation, data collection and data analysis. Prada RFID chip. Jinfeng Company collects data and turns it into a service provider. Exhaustion method improves search quality. Google autopilot.
New technology+original industry = new industry. Steam engine, textile, transportation, ceramics. Electricity, elevators, transportation and communication. Computers, banks, securities dealers, agriculture.
Big data, machine intelligence. IT industry chain. After the industrial revolution, promotion is needed, followed by advertising, business chain and service industry.
GE's smart refrigerator.
From the perspective of industrial development history, new technologies will inevitably bring about industry subversion. Embrace new data and realize transformation with the help of new technological revolution.
Technical challenge
Collection, storage, transmission, processing and application.
Computers, sensors. Moore's law. Parallel processing of big data
You don't enjoy it if you have data, and ratings are a problem. Nest intelligent air conditioning controller, deliberately collected information will be deformed.
The amount of data generated is greater than the amount of data that can be stored. Data security, integrity and loss prevention. Feel free to retrieve questions and access calls.
Parallel computing and real-time processing. Data mining noise. Early warning of abnormal operation with big data. Privacy issues, insurance companies, black-hearted merchants, plane tickets.
The related technologies of big data are relatively mature.
Israel's smart agriculture, a warrior trained with big data. Industry 4.0, Tesla robot, genetic testing, personalized medicine. Medical image detection, Da Vinci operating table, IBM Watson. Pharmaceutical revolution, cancer treatment, chemotherapy. Case analysis of lawyers, automatic writing of press releases,
Lawyers, doctors, finance and other industries that need high IQ can be replaced by machine intelligence, basic assembly production, Foxconn and so on. Will have an impact on the whole society.
Intelligent revolution
People flow forecast, intelligent transportation, intelligent signal lamp, intelligent route. Identification, counter-terrorism. Bitcoin, blockchain. Track every transaction. Personalized medical care. A society without privacy.
Agriculture to industry, industry to electrical appliances, workers from farmers to workers, and then to the service industry. What about the unemployed? British global colonization, American western development, World War I in Germany. At first, the new technology benefited very few people, and most people were eliminated. It took half a century or two to benefit the whole society, and it will take longer to spread to the whole world. General retirement benefits, people at the bottom can't find jobs.
Strive to be a 2% person, a new way of thinking, and actively embrace big data and machine intelligence.
Reading experience in the intelligent age 6
The content of this book is relatively shallow, basically belonging to the kind of telling stories and browsing. The author's writing, knowledge and experience are excellent, which is not reflected in this book.
The only bright spot is that I didn't know much about the specific relationship between big data and artificial intelligence before. Explain this book briefly.
Traditionally, machine intelligence means that machines can think and solve problems like people. However, because humans have not learned to solve problems in the current scene, they often apply other knowledge to solve this problem, so it is difficult for machines to learn "human" thinking in terms of migration ability. So what is the relationship between big data and machine intelligence? Another way to develop machine intelligence is to completely abandon the road of "thinking like human beings" and directly aim at the ultimate goal of "solving problems". For example, in translation, the previous idea was to let the machine understand all kinds of grammar. Now it is not necessary for the machine to understand grammar, but to "learn" hundreds of thousands of sentences directly, and the dosage changes to achieve qualitative changes. In other words, when you see more and more expressions, the accuracy of translation will continue to improve.
Reading experience in the intelligent age 7
Generally speaking, the main content of this book is to discuss how machine intelligence based on big data develops, how it will affect all aspects of people's lives, and how to deal with it.
Big data has brought profound changes to the world, the most profound is the change of thinking mode-from causal mechanical thinking to relevant thinking. What has changed here is not only the means to solve the problem-now we have the ability to analyze the problem in the complete works, and finally we have the ability to accurately describe the outline of the group on the basis of the complete works, as well as the way to look at the problem-we are no longer "careful assumptions and bold verification", but "reducing assumptions and self-evident data". This transformation actually makes machine intelligence a foundation for becoming a public facility-it is not necessary for everyone to understand how machine intelligence works, but for users, it is only necessary to understand that there is such a tool to find appropriate conclusions in a complete data set. Therefore, this will quickly trigger a commercial revolution.
In fact, this revolution has already begun: in the book, it has been listed that traditional industries, including bar and refrigerator manufacturers, use big data to improve their business conditions. However, there are still more industries exploring how to effectively integrate data and explore new models-first, the problem of data acquisition, and finding an effective way to obtain data, which may not be just nine Niu Yi hairs; Secondly, how to find value from these data still needs a guiding direction; Finally, how to use the rules of data discovery to design a more reasonable model.
Another wonderful content in the book is the last chapter, which discusses the positive and negative effects brought by the intelligent revolution. I think the author actually asked such a question: the intelligent revolution has greatly liberated the productive forces and promoted production, and only fewer people are needed to feed the world. What about the rest? By discussing the historical process of the first, second and third industrial revolutions, the author draws a conclusion that it is not optimistic-only through the role of time can the great liberation of productive forces brought about by the revolution be released. Unfortunately, we are in this historical process, and we are fortunate to be in this historical process.
So the whole book can be said to be a science book. After reading the whole book, you can further understand the most fashionable concepts-big data, machine intelligence, blockchain and so on. , especially big data. But I think it is more like a propaganda book-because it depicts two worlds, one is a world that participates in the revolutionary process and keeps up with the wheel of history, and the other is a world that resists the revolution and is abandoned by history-and the difference between the two cannot be ignored. As for whether it will really be so serious, this is a matter of opinion, but history tells us that at least the first three times were like this. If you have doubts, you'd better pray that there will be something new under the sun.
As for the revolution, there is no doubt about it. Of course, this is known before reading a book. In fact, many people already know that it is difficult not to think about the impact of big data on the world in an environment where the whole society is discussing big data. The utility of this book is only to let readers know how irresistible this process is.
Reading experience in the intelligent age 8
Artificial intelligence is a promising field in the future 10, and it is also our career vision in the data industry. After reading my book, I am more determined. This is a predictable and interesting direction. The following is my understanding and summary of data analysis and application in the process of reading this book.
Basic idea
Classical thinking: at present, when doing data analysis, traditional logical reasoning analysis thinking is adopted. Ask questions first, and then verify them through rigorous logical reasoning to explain business problems.
New ideas: try to use correlation to analyze data mining; It is some technologies of data mining, such as clustering, decision tree, random forest and other advanced statistical models. This kind of thinking is generally a high-tech threshold problem that classical thinking can't solve, and it is also in line with the current popular big data thinking and artificial intelligence thinking. The logic of the new way of thinking is to start with correlation analysis, find out the related factors that lead to the problem, and then explain the business logic behind it.
application area
Classic thinking: in line with the thinking mode of the human brain, from-> a; b—& gt； The logical order of C. It is very efficient in solving small and beautiful independent cases. For example, the question "Why did the conversion rate of a certain page suddenly rise today?" Through logical reasoning, drilling down step by step can quickly locate the cause.
Reading experience in the intelligent age 9
As students in the new era, we should not only compete with our peers now. And compete with some robots in tin. You may think it's ridiculous, but this is the trend of social development.
In all kinds of platform hills, you can see some variety shows in which robots compete with people more or less. After the robot defeated human beings in Go, it began to learn to write poems again! You say, as a student of the new era, how can you not be scared? Hawking once said: "The rise of powerful artificial intelligence is either the best thing or the worst thing in human history."
Now all industries have been replaced by robots. In inland developed areas, robots have occupied all industries, such as banks, and are used less. They are rapidly spreading all over the country. I'm afraid we won't need humans to farm in the future. What can humans do? In such a new era, the employment problem gives people great pressure. How many people will lose their jobs in the banking industry alone? The development of intelligent technology should also make us think about the pros and cons.
Judging from the current situation, the continuous development of science and technology. It is not impossible for robots to rule the earth. What can humans do when all jobs are replaced by robots? What about the value of human existence?
In such fierce competition, only hard work can make you stand out.
Reading experience in the intelligent age 10
We live in an era of advanced technology, and all kinds of artificial intelligence products appear in our lives, such as parents' smart phones, intelligent temperature measurement sheds at the school gate, face recognition systems at the entrances and exits of each community, intelligent sweeping robots at home, automatic checkout systems in supermarkets and so on. These smart products have brought a lot of convenience to our lives.
On the road of science and technology, intelligent human beings are still creating. In the future, there may be intelligent robots, smart houses and smart cars ... I am particularly looking forward to getting on the smart car and experiencing the different feelings of driverless driving. At this time, the car can accurately stop in front of me according to my instructions, safely send me to my destination, and then automatically identify the parking space and park it. Wow, I think it's cool!
I looked up a lot of information about smart cars on the Internet, and learned that driverless driving is the development direction of future cars, and it is also a very complex high technology, which needs to rely on artificial intelligence, visual computing, radar, monitoring equipment, global positioning system and so on. To put it bluntly, it is to equip a car with a smart brain, keen eyes and flexible hands and feet, so that it can become an "autobot", drive more safely and intelligently, and better serve mankind.
As a successor in the future, I will study hard and contribute to the future high technology.