He Fan: What will the future factory look like after the machine replaces people?
Text / Sina Finance opinion leader column (WeChat public number kopleader) columnist He Fan
There are two ways to replace people with machines. First, robots have self-awareness and will not listen to human command. Second, people and machines are integrated.
He Fan: What will the future factory look like after the machine replaces people?
Business management consultant Warren Bennis told a joke. He said that there is only one person in the future factory, a dog. People are going to feed the dog. The dog is going to look at the person and not let him touch the machine.
One day, robots will replace people. If a prediction is made about species evolution, then the next species after humans should be a robot.
It is recommended that you read a book on "How to Think about Thinking Machines", which brings together the top experts from all over the world to view artificial intelligence. Of course, even experts are mostly speculative about artificial intelligence. No one knows what will happen in the future. If you simply say, there may be two paths in the future.
The first path is that robots completely replace people.
Robots not only learn the human mindset, but also do better than humans. Robots also learn human emotions and are more rational than humans. The robot has self-awareness and will no longer listen to human command. This is not impossible. From the root to the source, people's thinking and emotions are nothing but physical and chemical reactions, but we know very little about their principles.
The second path is the integration of people and robots.
Mobile phones allow us to become "clairs" and "shun the wind", and can communicate with others in real time and without geographical restrictions. Big data makes it easier for us to learn and communicate better. People use a variety of artificial organs. In the future, people are likely to use more technology to improve our memory, extend our life and regulate our emotions.
· Bad memory? Just connect a USB flash drive.
· Character temper? I took a piece of medicine and changed it.
· Want to experience the Antarctic expedition? You can buy a personal memory from someone else.
· Can't understand Javanese? The machine will help you translate directly.
No matter what happens, we can imagine the final result: the human species will be completely changed. Life and work will be very different from the past, and even human survival will encounter challenges.
How does artificial intelligence appear? This is a very complicated issue.
We may wish to understand from the simplest point of view. In the past, computers were run by "programs". The programmers imagined various situations that might occur, and then told the machine how to deal with a situation.
This will bring a challenge. If it is a very complicated problem, there are many links, the machine must exhaust all possibilities at every link, then the complexity of calculation and judgment will grow exponentially until the machine Completely collapsed.
People also encounter various complicated problems every day. Whether to marry or to have children is an extremely complicated issue. There are endless possibilities in life, and there are all kinds of contingency. How do people deal with complex problems?
Our way of thinking is lazy. For example, in the first pass, there are two choices, we will choose one, for example, we choose A. Going forward, I met the second pass, and there are two more choices. Let's choose one more, for example, we choose A1. If A1 is a dead end? We quickly returned to the nearest branch point and chose A2.
Loop back and forth until you find the right path. This way of thinking looks stupid. Isn't this just a matter of luck? This is true. This is because people's memory storage capacity and computing power are seriously insufficient, and they come up with a way to make it happen, but it turns out that this is the only correct way to solve complex problems.
Our most intuitive experience with intelligent robots may be the home sweeping robot, a guy called Roomba.
When designing Roomba first, the designer was very upset. The rooms in each family are different. Some people live in villas, some live in a dim room, some rooms are square and some rooms are extremely irregular. If you want to pre-enter all the floor plans, it is almost impossible. After changing my mind, it suddenly became clear. Roomba's design philosophy is to let the robot learn by itself.
When Roomba first arrived at your home, it would be like drunkenness, hitting the wall everywhere. In fact, it is learning. It doesn't matter if you touch the wall, it will treat this as a failed attempt and record the result. As long as it records every failure and continually corrects it, it will become more and more proficient. In the end, it is like a puppy that is so happy in your home, coming and going, and comfortable.
In a nutshell, this design idea is the "trial and error method." Machine learning is a process in which computer algorithms continuously improve themselves in analysis and prediction.
The methodology of cognitive robots is nothing more than the most basic probability theory, but its technological advancement is that cognitive robots have begun to understand more complex unstructured information. In other words, robots can not only understand numbers as they have in the past, but also "understand" images, "understand" people's speeches, and so on.
Take machine translation as an example. IBM set up a team when developing translation software. At first, they hired many linguists, hoping that linguists could teach different grammars of machines and then let machines learn various languages ​​based on grammar. Later, they found that this would not work. The easiest and most rude way is to put a lot of linguistic data into the computer and let the computer go "trial and error."
In the beginning, the translation of the computer must be nondescript, but slowly, if you give the computer enough correct and wrong examples, it will slowly figure out which statements are not authentic and which are more authentic. It's getting faster and faster. It can learn Chinese, Russian, Bantu, and Nepali in the same way: in fact, it is not learning foreign languages, but processing statistical data. In the future, we will probably not have to learn foreign languages ​​anymore, and artificial intelligence will do better than us. It can master a variety of languages, relying on big data and "trial and error".
Speaking of artificial intelligence, we often have a fear that our work will soon be replaced by machines. Indeed, more and more work will be replaced by machines, but it is still very early from the era when I predicted that the robot replaced humans. We are still in the early stages of artificial intelligence.
Today's artificial intelligence is mostly limited to a specific area. Roomba is responsible for sweeping the floor. Some artificial intelligence is to translate language or help doctors diagnose diseases. They all have a division of labor. In their respective fields, they are entirely possible to replace many routine human work, but will robots suddenly become omnipotent? Will your room's Roomba sweep away the smell one day, and decide that you don't want to sweep the floor, you have to design the car? At least for now, this possibility is zero.
Task automation and job automation are two different things. Job automation means that the machine completely replaces people, and task automation does not steal human work.
For example, due to the industrial revolution, the textile industry changed from a handicraft industry to a modern industry in the 19th century. 98% of the labor in the textile industry has been automated, so is the number of employed people in the textile industry correspondingly reduced by 98%?
No. The number of employed people in the textile industry has increased. This is because after the productivity is greatly increased, the price of the product will decrease and the demand for the product will increase. In the past, many people only had one set of clothes, and they were hand-stitched by mothers. In the past, it was "the line of the mother-in-law, the body of the wanderer." Now, the wanderers are wearing the clothes they bought, and everyone’s wardrobe is full. In addition, the demand for curtains, carpets, sofa covers, and a variety of textiles has increased significantly, so the increase in demand offsets the machine's replacement of labor.
The same story is still happening today. Automated teller machines began to appear after the 1990s. At the beginning, people thought that with an ATM, they no longer needed a bank teller. What happened?
The number of bank tellers has increased, and it has increased faster than the entire US labor market.
Barcodes have been around since the 1980s. Scanning the barcode can reduce the checkout time of the cashier by 18-19%, but the number of cashiers increases.
Since the late 1990s, law firms have increasingly used electronic document retrieval software, which is the work of paralegals, but the number of paralegals has grown rapidly.
We have all heard that technology can create new job requirements, such as data science engineers. But at the same time, technology can change a lot of traditional routine work.
Bank tellers do not need to pay in cash, they can spend more time helping customers deal with more complex matters.
Lawyers no longer have to find documents in the archives, they can help lawyers better maintain customer relationships.
The computer that sees the CT film does not completely replace the doctor's work, and the doctor can use the computer to further improve the quality of the diagnosis.
There is no substitute for the designer's work in various design software. On the contrary, more people will be able to enter the business more easily.
So, the good news is: In the future factory, there will be a machine, a person, a dog. Some jobs are people to do the Lord, machine assistance; some jobs are machine owners, people to assist. As for the dog, it was quietly lying there, thinking like a philosopher.
(The author of this article: Chief Economist, Chongyang Financial Research Institute, Renmin University of China)
10 digit calculator
10 Digit Calculator,Sharp 10 Digit Financial Calculator,10 Digit Scientific Calculator,Electronic Scientific Calculator 10 Digits
Guangdong Shunde Zuncheng Electronic Technology Co., Ltd. , https://www.junno-electronic.com