Reading as a god

Chapter 241
"Deep learning is a kind of learning. It uses nonlinear elements for learning. This kind of nonlinearity is difficult to analyze accurately from a mathematical point of view, so we now have the opportunity. We may need to really apply some standards and methods, that is, physics. learning methods to achieve progress in artificial intelligence.

I think this is a gratifying development, and I hope that computer science can also acquire the depth and breadth of natural science. "

"So you also think we shouldn't ignore natural intelligence?"

"Yes, I think relying solely on artificial intelligence is not perfect. The future intelligence should be multiple intelligences, which not only has the shadow of artificial intelligence, but also includes natural intelligence~"

Having said that, replicating natural intelligence into technology is not an easy task!

First of all, when it comes to natural intelligence, it will inevitably involve interdisciplinary subjects~
This will undoubtedly cause trouble, and people are always afraid of trouble.

The most terrible thing is that in most civilized societies, some of the skills of natural intelligence have almost become a luxury, a kind of reminiscence.

Advances in technology can satisfy most of our biological needs.

Compared with some other intelligences, natural intelligence is less "valuable" today than it was in the past.

Most of us are already alive and well without natural intelligence.

But of all the intelligences now defined, there is a certain degree of natural intelligence that is not meaningless to any individual.

Zhang Shan was surprised that Mr. Yao actually chatted with him about artificial intelligence for so long~
Yao Lao has always attached great importance to artificial intelligence.

Regarding the question of "the essence of artificial intelligence and when the theory of artificial intelligence will be born", he said: the development of artificial intelligence is inseparable from the foundation of the previous theory, but with the further exploration of technology, deep learning It may not be the ultimate theory to solve the problem. "

As for "in today's artificial intelligence practice, where is the next innovation point of basic theory", he believes: "Deep learning has opened a window for artificial intelligence and human beings on the issue of 'what is intelligence'. Although it is still It is not possible for human beings to solve this kind of problem completely, but some hints are given. This is what makes people's eyes shine, and it also makes our research more motivated. But from a basic theory point of view, how to use interdisciplinary Finding the mystery of artificial intelligence with a method is the most challenging problem that people are facing at present.” The exploration of artificial intelligence requires not just a computer scientist, but a scientist. "

It’s no wonder Mr. Yao attaches so much importance to it. Artificial intelligence has brought about changes in labor and production relations, and its impact is no less than a revolution!
Artificial intelligence is a new beginning for us. We need to understand the underlying things before doing algorithms, which means that 99% of things in the future will be done in new ways.We use the spirit of a physicist to design algorithms.

The present is the best era of artificial intelligence, and artificial intelligence research is in its most prosperous period, because this period gives a new generation of talents the opportunity to become the next Newton or Einstein.Now artificial intelligence is getting hotter and hotter, attracting more and more people to pour into it.Perhaps, everyone can calm down in the hustle and bustle of the environment and find out the essential innovation of artificial intelligence, which is the key to the next door.

Nevertheless, Zhang Shan passed the information through Jarvis (Yao QZ, male, Han nationality, native of Xiaogan, Hubei, born in Shanghai, Chinese computer scientist. In 2000, because of Yao QZ's understanding of computing theory, including pseudo-random number generation, cryptography For his many contributions to the complexity of learning and communication, the Association for Computing Machinery (ACM) awarded him the Turing Award of the year) Knowing that Yao Lao’s specialization is not in the field of artificial intelligence~
In order to deal with the possibility that Mr. Yao suddenly talked about his field of expertise, Zhang Shan has been supplementing the general ideas of various papers written by Mr. Yao through Jarvis~
It was totally unexpected that Yao Lao would suddenly pull him to chat about some topics that Zhang Shan didn't expect.

Good guy, is it completely useless to prepare so many drafts?

~~
At this moment, Yao Lao suddenly heard asking:
"I heard that you made an agreement with Professor Wu, the head of your department~"

Guys, do so many people know about this now?
For a while, Zhang Shan remembered the two "overlord" clauses of Professor Wu:
"Zhang Shan, you must prove that you are qualified for this major. If you really want to go to the Department of Mathematics, you must complete all the courses in the Department of Computer Science within one year, and your grades must be in the top 5% of the same department. By then, are you willing to go to the Department of Mathematics and Otherwise I won’t interfere~”

"...For you, you must publish a paper in the computer field, reaching the level that can be included by any one of SIGGRAPH, SIGCOMM, ACL, AAAI, and IJCAI."

Zhang Shan was about to tell the story in detail, but saw Yao Lao waving his hand:
"I have known your agreement for a long time. In fact, it is my attention that you must publish one paper in the field of computer science. I even planned that you should publish at least three papers at the beginning. It was Professor Wu who spoke for you before I agreed to publish only one paper." article~"

Zhang Shan:? ? ?
Sure enough, "a person grows into a fine man", Yao Lao, whom I have always admired, is actually the "culprit" who made me so tired now~
Is the old man who co-authored you the instigator?

You actually plan to ask me to write three papers, and I’m not an old hen. Could it be possible to mass-produce papers~
"Zhang Shan, have you paid attention to quantum computers?"

"I know that a quantum computer is a device that uses quantum logic for general-purpose computing."

Elder Yao went on to say:

"Yes, unlike electronic computers (or traditional computers), quantum computing uses qubits to store data, and it uses quantum algorithms to perform data operations.

One of the most important starting points for people to study quantum computers is to explore the computing limits of general-purpose computers.

When using a computer to simulate quantum phenomena, the amount of data becomes huge due to the huge Hilbert space.

The computing time required for a complete simulation becomes quite long, even unrealistically astronomical.At that time, Richard Feynman thought that if a computer composed of a quantum system was used to simulate quantum phenomena, the calculation time could be greatly reduced, thus the concept of quantum computer was born.Semiconductors record and calculate information by controlling integrated circuits, while quantum computers hope to control the state of atoms or small molecules, record and calculate information. "

"Also, Zhang Shan, do you understand pseudo-randomness~"

After preparing the draft for a long time, Zhang Shan almost blurted out:

"Pseudorandomness is a process that appears to be random, but is not.

For example, a pseudo-random number is a seemingly random number sequence calculated using a deterministic algorithm, so the pseudo-random number is not actually random.

When calculating the pseudo-random number, if the starting value used remains unchanged, the number sequence of the pseudo-random number will also remain unchanged.

The randomness of a pseudo-random number can be measured by its statistical properties, the main features of which are the probability of each number appearing and its relationship with other numbers in the number sequence when it appears.

The advantage of the pseudo-random number is that its calculation is relatively simple, and it is difficult to deduce the algorithm for calculating it using only a few values.

Usually people use a pseudo-random number, such as the time on the computer, as the starting value for calculating the pseudo-random number..."

Yao Lao looked admiringly: "That's right, if you have achieved something in mathematics, you will look at these computer problems more deeply~
You may think that it is too much for me to ask you to make such a result in one year, but it is not too easy for a person who can publish world-class papers in the field of discrete mathematics!

You should also know that discrete mathematics and computers have almost an inseparable relationship...

In addition, you are very concerned about cutting-edge fields such as quantum computing, artificial intelligence, and cloud computing, so I have strict requirements for you!

In fact, three papers are not index work for you at all, for you, such work is completely at your fingertips~"

Zhang Shan:. . .

I always feel that this logic is weird. According to this logic, I am tall and have good physical fitness. Wouldn’t it be good to play basketball?

Zhang Shan thought in his heart, the fluent answer was just that he had already prepared through Jarvis.

It doesn't matter how you think, but some respect is essential:

"Thank you Yao Laoxin, I will definitely work hard to complete the task~"

"By the way, Mr. Yao, can you tell me about your own experience~"

"I, when I was 21, went to graduate school in the United States, and then I returned to China to work at Shuimu University.

The purpose at that time was to promote the development of computer science.In fact, what I was thinking at the time was to go to an elite university like Mizuki. One of the most effective things is to produce elite computer science talents. They should be at the same level as those produced by first-class universities in the West.

For us to have such a program, it is not enough for us to just teach, we must have a research environment to support our work.

Later, I finally started to set up the Institute of Computer Theoretical Science in Tsinghua University and conduct my research.Overall support for educational programs.

This is a very difficult job. We need to produce some elite students, especially elite students at the doctoral level.

In developed countries, it is almost impossible, but China has some special advantages, such as the well-known Chinese students are very good.

I have this ambitious goal, but it is very difficult to achieve it.Because only when these students are really convinced by us, and they believe that the conditions we provide are comparable to the conditions that the West can provide them, they will stay.You must also convince the government and various institutions that you have such a goal, and you must obtain their financial support before you can carry out this project.That's what I've been doing for the past four years.

The educational program I'm working on now includes an elite course for undergraduates, and we have a special pilot course that admits 30 students every year from the new first-year computer science students.

At that time, there were more than 800 computer science students in Tsinghua University. We recruited 30 new students in total. We will develop ten special courses for them. These ten fields are what we think Chinese students are in great need of, including computer science. Theoretical science, that's where my expertise lies.

There is also the field of computer systems.

I don't have any expertise in the field of computer systems, but I am very lucky to have a few very good Microsoft researchers who can help me in this area.

The second project I started later was a postgraduate project, which was in my field of expertise.

The previous project was not a course of specialization, it was just a computer science.Our current postgraduate project is in computer research theoretical science, which requires some international cooperation and exchanges. With this course, there is also this demand. We don’t have enough teachers now. The most important point is that this is also the fastest way. , allowing us to achieve first-class results.

As I said just now, when we play Go, you are often at the same level as your opponent.So, if we have enough international interaction, these very talented Chinese students, for their own self-esteem, will be as good as their counterparts.

As part of this program, we give our students at least one year of public funding to study abroad, and we pay for it, and that's one of the selling points we have to attract students to our graduate programs, because I can ask them, do you have Confidence to be admitted to Stanford University.They will say no.I just told them, if you are in one year or two years, I will send you to these universities, no matter which major you want to study, I will make these arrangements for you personally, so far these methods are very useful .

It is not enough to have the course itself, but also the whole environment.

Therefore, some senior students and graduate students can further understand the relevant research culture during the macro design process.

In China, at least in terms of computer science research, our experience tells us that it is very difficult to attract Chinese professors working in the West to come back to work in China, and it may be more difficult than retaining physics and biology research talents.So, I later decided to build a very strong group, which is the postdoctoral supervisor.This group is composed of postdoctoral supervisors working in the West, including more than 20 well-known researchers.Some are also good friends of mine who have made a personal commitment to provide relevant support and mentorship to students.

Recently, I have taken a new step. As Professor Alferov once pointed out, sometimes it is never too early to start such an educational project even in high school. The earlier you start doing such an educational project, the better.

We’ve also launched some innovative new projects…”

Zhang Shan regretted asking such a question!
What Zhang Shan wants to know is Yao Lao’s growth experience, but Yao Lao’s talk is all about the experience of cultivating talents~
A person who has no selfishness~
This is the peach and plum without saying anything!
(End of this chapter)