The next 5 to 10 years will be a critical period for a new round of global scientific and technological revolutions and industrial changes, which will go from a buildup to a group burst. With the rapid development of the new round of global scientific and technological revolution, disruptive technological innovations are on the rise, including the most notable ones such as quantum computing, artificial intelligence and blockchain. Do these disruptive technologies have any connection with traditional Chinese culture? How are they related to basic sciences (e.g., mathematics and physics)? How to recognize these cutting-edge technologies objectively? This issue features a report by Zhang Shousheng, an American-born theoretical physicist and foreign member of the Chinese Academy of Sciences, who was honored in January 2018 with the People's Republic of China*** and the State Prize for International Cooperation in Science and Technology.

Currently, quantum computing, artificial intelligence and blockchain are the three most important basic technologies in the entire information technology industry. In the future, in order for information technology to truly be able to leapfrog development, it is necessary to pay attention to basic science, both physics and mathematics are needed, because physics and mathematics are closely related to the information technology revolution.

The discovery of the "angel particle" changed the dilemma of quantum computer development

Before we talk about quantum computing, let's tell the story of the scientific discovery related to the "angel particle". Many of the most interesting scientific discoveries of the modern era have been linked to changes in philosophical concepts, including those y rooted in the ancient philosophies of the Chinese people. This opposing worldview is also presented in the physical world of elementary particles.

Historically, there was a very great theoretical physicist, Dirac, who unified Einstein's special relativity theory and quantum mechanics, and in the process of unification he did a very simple mathematical operation, opening a root sign. When opening a root sign, there are always two solutions, positive and negative, and the general public may only care about the "positive solution", not the "negative solution". Dirac interpreted the "negative solution" to mean that all particles must have antiparticles, and predicted that all particles must have antiparticles.

In 1928, when no antiparticles had been discovered in the physical world, he was very much challenged, saying that his equations must be wrong. He insisted that his equations were right. After five years, he was very lucky, and sure enough, inside the rays of cosmic radiation, physicists found the antiparticle of the electron, which is the positive particle, named Dirac Sea.

Since then, elementary particle physics has had protons to find the antiproton, and neutrons to find the antineutron, and has been applied. For example, positron has been widely used in the medical field, there is a medical test called PET, the use of positron and negative electrons can be imaged to measure Alzheimer's disease, the best way is to do PET.

Today, the Chinese people are very concerned about the development of science. What is the biggest driver of scientific development? I think it is curiosity about life. Theoretical physicists in history, such as Isaac Newton, were under an apple tree, and gravity was discovered when the apple fell and inspired him. Einstein was riding in an elevator and felt that the up and down of the elevator was very similar to the action of gravity, thus creating the great theory of general relativity.

Also, science should be developed without fetishizing authority. After Dirac became a very famous theoretical physicist, scientists were very convinced that there are particles in the world and there must be antiparticles. But another great theoretical physicist, Majorana, he asked out of curiosity whether there could be particles in the world that did not have antiparticles. He invented the Majorana equation, which marvelously depicts that there is a particle that has no antiparticle, or it is itself its own antiparticle.

The entire physics community has since been searching for the coveted two particles, one called the "God particle," which was found in a European gas pedal in 2012, and for which the physicist who predicted it, Higgs, won a Nobel Prize, and the "Majorana fermion," which is a particle that has no antiparticle, but which has no antiparticle. ".

I work in theoretical physics, and the job of theoretical physicists is generally to make predictions and let experimental physicists test them. My experimental group predicted back in 2010 that Majorana fermions could be found inside a combinatorial device. But we still need to find a signal that can prove the existence of such particles.

One day I thought that the Majorana particle has only one side and no opposite side, so in a sense it's half the usual particle. Our theory group made the bold prediction that since the Majorana particle is not like the usual particles, in some sense it is only half of the usual particles. So it will have a different conductivity, and since the usual particles have conductivities that are integer multiples of 0, 1, 2, and 3, it will inevitably lead to half-integer multiples of the conductivity step. We predicted it would have a 0.5 or 1/2 step. Then our theoretical group worked closely with the experimental group and made experimental observations, and indeed, at 0.5, as you can see from the original pattern of the experiment, there was a step at 0.5, which proved the existence of Majorana fermions. We named it the "Angel Particle", and people really liked that name.

What does the Angel Particle have to do with the development of information technology?

Nowadays, computers have been divided into two categories, classical computers and quantum computers. Some problems can be easily solved by classical computers, such as multiplying two large numbers, which can be calculated very quickly. But a number to see if it can be split into the product of two other numbers, for example, 15 can be written as 3 times 5, this number is relatively small, you can calculate it yourself. But give you a very large number, the classic computer to calculate this number is not the product of two numbers need to spend a long time, because it uses the algorithm is the exhaustive method, all the numbers that can be divided one by one, and finally to confirm that this is not the product of two numbers, the classic computer calculations are very slow.

Classical computers can only use the exhaustive method to finally figure out an answer. But the quantum world is a very amazing world, a parallel world. For example, a famous experiment, if I put out a particle, like a photon, it has two holes, either the left or the right. But the quantum world has a kind of intrinsic parallelism in it, where an elementary particle passes through both holes at the same time at a certain instant. The image is not the image shown if it's either left or right.

The quantum world is inherently parallel. If you use the quantum world to do calculations you can do them in seconds, you can calculate all the possibilities at once, because the quantum world has its own true parallelism, and that's the most basic concept of quantum computing. But to really build this quantum computer is very difficult, for example, the most basic unit, the most basic unit of the classical computer is the bit, that is, the information should be either 0 or 1, with 0, 1 will be able to express all the information, this is the concept of the classical computer. But in the quantum world, a particle passes through the left hole and the right hole at the same time, in a kind of superposition state. A quantum bit can't be said to be a 0 or a 1, it's in a superposition of 0s and 1s. We can hear an analogy, Schr?dinger's cat is in a superposition of dead and alive. This is a very marvelous phenomenon. But because of this fundamental phenomenon, it means that a quantum bit is inherently unstable, and if you look around you can see that it's either on the left or on the right, it's either a 0 or a 1, and any noise will interfere with the quantum bit.

Recently, quantum computers have become something of particular concern to famous companies around the world and in the U.S. Google, Microsoft, IBM, Intel are all making investments, but fundamentally cannot solve the problem because a quantum bit is very unstable, if one day tell us to do 50 quantum bits, but the key question is how many useful bits, if there is only one useful bit, often in the This framework of quantum computing requires 10, 20 or even 40 or 50 error-correcting bits to serve it, making it difficult to actually realize quantum computing.

But the discovery of the angel particle has fundamentally changed the dilemma of quantum computer research and development, and this is a quantitative to qualitative change. The quantum bit itself comes with the ability to correct errors, which is what I call the ability to split the usual one quantum bit into two angel particles. Usually particles have two sides, angel particles only have one side, so angel particles are usually only half of a particle. So usually a quantum bit can be stored with two angel particles to store it. Once two particles are used to store it, they're in a distant place, they're entangled with each other. Noise in the classical world, they're not entangled with each other, so there's no way to destroy the quantum stored by the angel particles with noise, so that's a revolutionary change.

So I was speaking at the American Physical Society not long ago, and I said that angel particles are an exciting discovery, and that they can be used to make quantum computers with as many bits as they need, without additional error-correcting bits, and with error-correcting capabilities, and that this is going to have a dramatic effect on the development of quantum computers.

Robots will be able to make scientific discoveries one day, and that day intelligent machines will surpass people

Artificial intelligence, as a basic concept, was introduced in the 1960s. The fact that AI is able to grow by leaps and bounds today is largely due to the aggregation of many new technologies. According to the iteration of Moore's Law, it is able to double every 18 months, and if quantum computing is used, it is not just doubled in 18 months according to Moore's Law, but completely changed from quantitative to qualitative. Over the years, human computing power has grown. The birth of the Internet and the Internet of Things have generated massive amounts of data. Intelligent algorithms have changed by leaps and bounds. Big data can help machines learn. However, the basis of artificial intelligence is all kinds of data, even the best algorithms, even more powerful computers can not become artificial intelligence without data.

Artificial Intelligence, although we now see that it is making rapid progress, but I think it is still in the very early days. Why do I say this? Do a simple analogy, for example, we once saw the bird fly, people also want to fly, but early learning to fly is just simple bionic, in the human arm tied to the wings, which is simple bionic, but really reach the realm of flight is due to mankind's understanding of the first principles of flight - aerodynamics, physical principles and mathematical equations can be artificially designed for optimal use of artificial intelligence, but also to achieve a better understanding of the realm of flight. With the physical principles and mathematical equations it was then possible to artificially design the optimal flying machine, and now airplanes fly high and fast and well, but not like birds, which is a very central point.

Now AI is mostly simply mimicking human neurons, but what we should be thinking more about is that there is an opportunity for a major breakthrough in basic science here, and we need to really understand the fundamentals of that wisdom and intelligence so that we can really make a fundamental change in AI.

What kind of basis can be used to really measure AI to human standards? Some people may have heard of the Turing test, which says that a person talks to a machine but doesn't know whether the other person is a person or a machine. The whole conversation is a sign that the robot seems to have reached the human level if you spend a day not feeling it at all. While Turing was a great computer scientist, I don't agree with this method of judgment. Many human emotions are not rational emotions, and it may not be so easy to get a rational machine to learn an irrational human brain.

So I would like to propose a new judgment method, intelligent robots one day really have beyond human intelligence? I think one of the greatest things about human beings is that we are able to make scientific discoveries, and the day that robots are really able to make scientific discoveries, that's the day that machines will surpass human beings.

Recently, I wrote an article on artificial intelligence, which will be published in the Journal of the American Academy of Sciences, and it will mention that the greatest scientific discoveries of mankind are the theory of relativity, quantum mechanics, and so on, and the greatest discovery in chemistry is the discovery of the periodic table of elements. Can intelligent machines automatically discover the periodic table of elements without any tutoring? Can it help human beings discover new drugs, and can it discover new materials by using machine learning? These are the criteria for judging the level of artificial intelligence.

To realize the development of blockchain and AI mutual ****existence, they will be the most valuable

In today's world, individuals will produce a lot of data, personal genetic data, medical data, education data, behavioral data, etc., which is particularly needed for the development of artificial intelligence. A lot of data is held inside central institutions, without achieving true decentralization. The creation of blockchain can produce a decentralized data market.

I describe the whole idea of blockchain in a phrase called "In Math we trust", which is based on math. The most basic thing in the whole blockchain and the whole information technology field is the basic math, which is able to protect the privacy of individuals in the data market, but also able to make reasonable statistical calculations. For example, there's an amazing calculation called zero-knowledge proof, which is able to prove to you that my data is very valuable, but doesn't tell you where the really private data is.

With blockchain, data markets can make society fairer. The biggest unfairness in modern society is that people tend to discriminate against some minorities. But what is needed most in the process of machine learning is the data owned by those minorities. If today's machine learning has reached 90% accuracy, so that 90% to 99%, it needs not the data that has been learned, but the data that is different from before. It's often the minority of data that is most valuable for machine learning. Once we have our data building on top of the blockchain, and after we add these wonderful mathematical algorithms, we'll be able to have a benign data market. Within this world, reaching the idea that blockchain and AI **** each other to exist, they are going to be the most valuable.

The whole blockchain, the public's understanding of it is not yet the most fundamental first principles of understanding. In terms of the most basic principles of physics, to reach a **** knowledge is like everyone agrees on the same "ledger", equivalent to the physics, magnets are originally disordered, but to the ferromagnetic state inside the direction they point is the same.

The phenomenon of entropy reduction is also found in the natural world. The entropy of a state in which everyone is pointing in the same direction is much lower than the entropy of a chaotic state. Reaching this **** knowledge is very difficult because entropy is always increasing.

The systems that can reach a ****-sense on the blockchain all use an algorithm that requires energy consumption. This thing sounds unreasonable, why does the account consume energy, but from the second theorem of physics, this is a very reasonable thing, because to reach a **** knowledge itself is entropy reduction, but the entropy of the whole world must increase, so in reaching a **** knowledge at the same time must exclude some other entropy. This kind of mechanism without centralization is very similar to the natural world inside the magnet from the disordered state to reach the orderly ferromagnetic state, consume energy to pay the price is also the inevitable trend.

So the ideal information world, is the future of each person has all their own data, completely decentralized storage, so that hackers can not hack everyone's data. And then using some encryption algorithms on the blockchain to really be able to achieve both the protection of personal privacy, but also to be able to make good calculations, and not to happen like in Facebook where a lot of personal data is stolen.

The problems of quantum computing, artificial intelligence, and blockchain technology that we are trying to solve today are the problems of the whole human race, and Chinese scientists will face a very big opportunity. In addition to doing a good job in applied science and technology, there should also be truly original breakthroughs in basic science, such as the physical and mathematical principles described above, even though they sound abstract, such as the principle of entropy increase, and the positives and negatives of electrons. The wonder of the world is that basic science can provide the entire information technology industry with a broad new development prospects.