"My friend h is an old driver. On average, he travels hundreds of kilometers every day to pick up hundreds of thousands of people, and the price of his car is second to all luxury cars on the street - yes, he is a subway driver.
He is very unhappy. The driver industry is easy to sigh about life, especially for subway drivers who run the same route every day. He is tired, stressed and doesn't earn much. No longer young, he often thinks about changing his industry and life; However, because he is no longer young and because of the limitations of "subway driver", his life is becoming more and more fixed, and the changes he expects are becoming more and more difficult to happen.
I have some sympathy for him, not only because we have known him for many years, but also because his situation is actually the same as that of many people. In poor families, children often lack the right to choose. This means that before they really understand what they want to do and what they should do, his education planning and career planning have locked him in "what he can do". Only a few strong and lucky people can break the shackles of "have to" and realize a "I can" future.
He studied design, UI and programming. However, lack of experience and over age led most companies to decline his job search. What's more terrible is not the embarrassing situation, but it's because he said to me: in fact, I don't like the things I learned.
He sighed to me: if I had known that all this would have such a great impact, when I went to college, I would have chosen a major that was easy to find a job.
Intelligent energy
Since Obama appointed artificial intelligence as an important development direction in the White House, the world has become more tolerant of artificial intelligence. Many years ago, we were still imagining the plot in the terminator. People worried about the world and were afraid that robots would dominate mankind; Many years later, the fear that robots make us unemployed is close at hand, but we selectively ignore the world 3 or 5 years later and begin to praise the miracle of alphago.
In the eyes of optimistic seekers, the development of artificial intelligence can't stop for a moment. AI is really great. They are far faster and more accurate than people. They don't complain and don't sleep. They connect to the latest news as quickly as you smile when you receive your salary. They will be the best substitute for the disappearing cheap labor force. As for where the replaced cheap labor force should go, many people believe that the huge improvement of social efficiency will bring common prosperity. At that time, unemployment will not be a problem.
Hearing this, many people feel relieved, especially those who are not familiar with artificial intelligence. They may not seriously think about whether unemployment is a problem. Another circuit in their brain was triggered first: "it's not me who lost my job."
Isn't it really you?
I don't know which industries have such confidence. If you know a little more about artificial intelligence, you will find that the future in three or five years is far from so optimistic. You may think that the mechanical arm replaces the assembly workers, you may think that high-speed computing replaces the practitioners of Finance and finance, and you may think that deep learning has crushed the traditional media and consulting. However, Professor Yang Yang of Ipin (former associate professor of Harbin Institute of technology and doctor of information in the United States) said: you may underestimate the creativity of the machine.
Yes, machines are very creative. Let's take the simplest example. On the Chinese Internet, "automatic poetry writer" has a history of more than 20 years. Strictly speaking, poetry writers are best at producing all kinds of garbage, which accounts for 99.9% of their works; But the remaining 0.1% is not necessarily worse than the works of proficient poets.
Why? How come? In terms of probability, monkeys can also type handed down poems, but it takes a lot of monkeys and a lot of time. A lot of monkeys and a lot of time are the easiest things to achieve in front of AI; Even if the pass rate is only 0.1%, will it take a second to produce 1000 poems?
The director of an advertising company, instead of asking a lot of subordinates to write hard, come out 100 paragraphs in three days and pick out the right ones, why not let the machine write 1000 paragraphs in three seconds? If there is no suitable paragraph in this 1000 paragraph, why not spend 3 seconds?
The CTO of an Internet company, instead of asking products, designs and programs to make 28 prototypes day and night and give everyone a 24K monthly salary, why not discuss with the machine? The latter will not complain about overtime, nor will it cause real people to fight quickly.
Many people, do you really think you won't be challenged by AI?
hesitate about what move to make
Of course, mankind reserves the right to choose. We may no longer need basic producers - even if they are engaged in creative work - but we still need the one who picks out the gold.
Even if it's not gold. Because whether a thing is gold is essentially based on human subjective judgment. The creation of machines is very good, but you don't like it.
It is this willfulness that enables mankind to determine its destiny. The awakening of consciousness is far from being realized even in the human body (otherwise you can revive the corpse with complete physiological structure), and it is far away in the field of artificial intelligence.
Yang Yang said: alphago is a good thing. Its emergence makes ordinary people pay attention to artificial intelligence and makes the whole industry prosperous. You may be afraid of alphago, but alphago has two problems that make him unable to replace people: first, alphago can win and know how to go next, but if you want it to explain why it is so settled and how it won, it can only be silent; Second, it doesn't know why to win.
In short, if you set alphago as an AI fighting for failure, it will become a "smart fool" who can fail 100% through continuous learning.
Therefore, Yang Yang said: I prefer to call "artificial intelligence" machine intelligence.
Education, recruitment and law. These are the three fields that Yang Yang's Ipin is involved in. He uses these three fields to explain to me the ability and limitations of artificial intelligence.
The first is education. Ipin's representative product "perfect volunteer" was released at the end of 2015. This product helps high school students view and choose their favorite schools and majors. With the power of big data, AI calmly, objectively and slightly cruelly tells students the possible consequences of filling in a volunteer: with your current grades, you have only a 31% chance of being admitted to a major in Tsinghua University.
This is an example of the combination of human subjective initiative and powerful machine intelligence. Of course, it is also in view of the system design of the college entrance examination: good test and good report are undoubtedly the two legs of the college entrance examination strategy for most people. Are your teachers reliable, your parents reliable, or do you read all kinds of materials by yourself? In this kind of pure information collection and data analysis, artificial intelligence is certainly more stable than anyone.
But at the end of everything, how many points you can get depends on the students themselves. This operation of self-determination fate may be one of the best modes for human beings to coexist with "machine intelligence".
No regrets
Similarly, in terms of recruitment and law, the rationality of decision-making is also closely related to the subjective conditions of the parties and the objective environment of the society. Subway drivers are not easy to change jobs. On the one hand, their ability is really limited. On the other hand, even if there are companies willing to recruit him, with his own search ability, he can only miss it.
However, this judgment is very rough. The monotonous resume of a subway driver does not mean that he is monotonous, even if it is only one line; The rigid recruitment notice of the company also does not mean that people who do not strictly meet the requirements are incompetent. In the final analysis, the words on paper only express part of us, whether it is needs or capabilities. AI can be as strict and rational as professional HR, but combined with the information of the whole network rather than being limited by hard words, AI is enough to break the threshold of resume and "use talents regardless of style".
This exception is undoubtedly of great significance to job seekers and recruiters. It is also part of the social value that can be displayed by technologies such as "semantic analysis", "big data" and "artificial intelligence".
Isn't that the case in law? The experience of each lawyer is hard to say, and the complexity of the case may not be expressed by the writing level of the person involved - non intelligent recruitment websites can more or less describe the pale image and recruitment requirements of job seekers by listing entries and filling out forms, but if they use the same way to "find a lawyer", the efficiency may be more limited.
In the traditional sense, we mainly rely on human self-determination to deal with massive information that is difficult to quantify; Whether it's medical care, legal aid or venture capital. C2C's website can display the most basic information, but the lack of expression means that users' search and decision-making are as slow as a snail.
For this reason, although C2C website has become more and more mature, its efficiency is bound to be low. With the introduction of "intelligence" products after semantic analysis, artificial intelligence can find matching lawyers even if users do not have legal knowledge and good expression ability to describe "I hit someone with my car".
Yang Yang said that just as map software has produced different products in static environment, walking environment and vehicle environment, AI will also turn capabilities into various forms for different scenarios. A person can rely on big data when filling in the college entrance examination volunteer, and can rely on big data when recruiting. It is also possible in other stages of life.
Ipin may release an artificial intelligence product to help users make "career planning" in April this year. This product may be simple in the early stage, but Yang Yang said that it will be refined continuously in the future. The purpose of Ipin is to make artificial intelligence cover all aspects of life, from primary school to university, from marriage to grave.
We once said that "life is like chess", but go, which contains infinite wisdom, has been captured by alphago. In that case, it seems quite interesting to let the brave and infinitely possible AI challenge the chess game of life.
Artificial intelligence Roadmap
A person's desired life is self-determination, and after the user makes the corresponding value judgment, artificial intelligence will draw a road map for the user to "follow and achieve the goal".
"Many of us know what we want in the end, but we are often careless about how to design the path to our destination." Yu Minhong wrote so in the "map of life".
The reason why we are not careful in our life may be that the calculation covering a wide range of fields and a large amount of data is beyond the reach of human beings. However, isn't it just appropriate for machines to do this?
When it comes to data, for artificial intelligence, the source of data largely determines the "success or failure" of products. Yang Yang said that there are so many big data products in China, and the technology alone is not without differences, but the differences are not enough to open the gap at the commercial level; Another hidden worry behind technology is really not optimistic - the digitization of information in China is not very ideal, especially in some fields, such as medical treatment.
In this environment, a large number of products using suspicious data will inevitably get wrong results; If the output result is wrong, it will undoubtedly cause harm to users. Watson of IBM has cut into the medical field, which must be based on the considerable accumulation of medical data in the United States. Since Ipin has even begun to get involved in the "career planning" of users, it is even more necessary to be vigilant - misleading users in filling in volunteers is probably more harmful than the "Wei Zexi" incident.
Yang Yang said that in the final analysis, artificial intelligence serves people and society. Social value often represents demand, and demand will undoubtedly bring the possibility of business. Yang Yang said that although it sounds absurd, career planning will become a huge market in the future.
Life is like a box of chocolates. We shouldn't know what's next. However, after solving the problem of data volume and reliability, with the maturity of career planning products, we may be able to find out the taste of the whole box of chocolate in advance. At that time, can the user's life still be in his own hands? Or will a "sybila" system control everyone who is consciously immersed in it? This may not be known. All I know is that all this may come true in five years.
Yang Yang thought it was funny. He said, as an undergraduate majoring in optoelectronics, how can I start artificial intelligence? The Ipin partner on one side said: Yes, yes, my science is polymer materials
At the moment, I think of my friend h, who is a subway driver, and many ordinary people who have fallen into eternal hatred., Technology Zone
Ma Huateng proposed the "Tencent Dingdang ecological Partnership Plan" to build a digital ecological community
Can enterprises rely on artificial intelligence to interact with humans?
What elements of mobile phone photography have been changed by artificial intelligence?
After splitting windows, can Microsoft really return as the king?
Huichun Technology: show major breakthroughs in IOT and artificial intelligence, and push gesture recognition“
Our other product:
