Blogs

artificial intelligence can be described as the “star” of today’s society, and many countries have formulated artificial intelligence development plans. However, the development of artificial intelligence is not entirely a technical issue. Its changes to human society are multi-faceted and have far-reaching impacts.

What is artificial intelligence? Let’s make a little definition: artificial intelligence is the imitation of human intelligence and attempts to achieve certain tasks. It mainly includes three aspects:

The first is computational intelligence, which involves fast computation and memory storage capabilities. From the point of view of computer scientists, artificial intelligence is first and foremost computational behavior, that is, involving data, computing power and algorithms.

The second is perceptual intelligence, which involves the machine’s perceptual capabilities such as vision, hearing, touch, etc., that is, the machine can capture and analyze the surrounding environmental information through various types of sensors, and make a reasonable response and response according to the requirements after processing. reaction.

The third is cognitive intelligence, which refers to the ability of machines to think and solve problems independently. The current artificial intelligence mainly stays at the first and second levels. Cognitive intelligence involves deep semantic understanding, which is still very difficult to achieve.

It should be seen that artificial intelligence is not only a buzz in the scientific and technological world, but also a major issue that the social science community needs to be concerned about. I came up with a concept called “Salvator”. There is a “Leviathan” in political science, which is a giant beast in ancient times, and was later used by Hobbes to refer to the state, describing the state as omnipotent. I changed the L in “Leviathan” to S, which means science. Science is powerful, but we humans have to tame it and not let it trouble us. This requires more thinking and practice in multiple dimensions such as ethics and legal system.

When will the “singularity” of super artificial intelligence appear?

I often do surveys in speeches, for example, when I ask the audience about when they learned about artificial intelligence. Many people will talk about the showdown between AlphaGo and Lee Sedol. In fact, the first wave of artificial intelligence that most people understand is already the third wave worldwide.

The first wave occurred from 1950 to 1970, when the main job was computer scientists working on machine reasoning systems, while inventing early neural networks and expert systems. The theoretical school of this period is known as Symbolism.

The second wave appeared between 1980 and 2000. Concepts such as statistics, machine learning, and neural networks that we are talking about now have all been proposed at this stage. The dominant school of theory at this time was called connectionism.

The third wave was after 2006, mainly due to the promotion of big data. Google’s successful use of big data to predict the flu has drawn the attention of health authorities, an important example of how big data and artificial intelligence are closely linked. In this wave, artificial intelligence technology and applications have been greatly improved, and algorithms centered on neural networks have made breakthroughs.

Regarding artificial intelligence, there are three related concepts that need to be clarified: the first is weak artificial intelligence, the second is strong artificial intelligence, and the third is super artificial intelligence. Weak artificial intelligence is dedicated artificial intelligence, and it is difficult to directly use it in other scenarios. The ideal goal of many scientists now is strong artificial intelligence, such general artificial intelligence can be migrated to other application scenarios. Super artificial intelligence refers to intelligence that exceeds human beings. It does not exist yet. I also hope that it will never exist, otherwise it will subvert the meaning of human beings.

However, American scientist and inventor Kurzweil believes that general artificial intelligence may surpass humans in the 1930s or 1940s, and regards this point in time as a “singularity”. To this end, a highly innovative Singularity University has also been established.

Bostrom, a futurist at the University of Oxford in the United Kingdom, also believes that it is very possible that superintelligence will surpass humans in the future. He divided superintelligence into three forms: one is high-speed superintelligence, the other is collective superintelligence, and the third is high-quality superintelligence.

High-speed superintelligence is similar to, but faster than, the human brain. In Bostrom’s words, high-speed superintelligence can do everything human intelligence does, but much faster.

Collective superintelligence consists of a very large number of small intelligences, and the overall performance of this intelligence greatly exceeds existing cognitive systems in many domains of generality. Collective intelligence is best at solving problems that are divided into sub-problems, and solutions to each sub-problem can be found simultaneously and individually verified. Bostrom believes that the integration of collective superintelligence can be loose or tight, that is, it can form a unified agent.

High-quality superintelligence is as fast as the human brain, but the level of intelligence is qualitatively superior to that of humans. This high-quality intelligence is compared with human intelligence, just as human intelligence is compared with the intelligence of elephants, dolphins, and orangutans. Bostrom believes that future superintelligence can acquire a new set of cognitive modules, and through the construction of complex knowledge engineering, general intelligence can gain new advantages.

What other important related technologies are there in the fourth Industrial revolution?

Why is artificial intelligence so important? Because artificial intelligence represents the fourth industrial revolution.

The cause of the First World War, traditional textbooks often emphasize the contradictions between Britain and Germany over colonies. Britain is an established empire with vast colonies, while Germany is a new empire with fewer colonies, so there is fierce competition. This conclusion is true, but also ignores a deeper reason, that is, Britain and Germany are fighting for technological dominance.

Britain is the leader of the first industrial revolution, and Germany is more of the leader of the second industrial revolution. Britain cannot tolerate the rise of “upstarts” like Germany. On the surface, it is a competition for colonies, because the colonies represent the origin of raw materials and consumer markets, but in fact the two countries are fighting for the dominance of emerging science and technology.

In the process of competition between Britain and Germany, the biggest beneficiary is the United States. On the one hand, the United States participated vigorously in the Second Industrial Revolution; on the other hand, it did not engage in direct conflict with Britain and Germany.

The third industrial revolution is mainly the information technology revolution, and the leader is the United States. In the process of leading the third industrial revolution, the United States has encountered two powerful opponents: one is the Soviet Union and the other is Japan.

The Soviet Union has made an effective layout in the information industry, and has advantages in the military industry, cutting-edge technology and the reserve of mathematicians. However, due to reasons such as technology and industrial system, the Soviet Union failed to carry out effective civilian transformation of high-end military industries, so technological innovation lacked sufficient application support.

Japan began to participate in the information revolution in the 1960s, and progress has been very rapid. In 1985, Japanese companies and American companies changed their roles in the market share of semiconductor production. Japan became the first, the United States became the second, and the share of other countries was less than 10%.

Faced with this situation, the United States adopted a series of methods to combat the rise of Japan’s information industry. For example, through the “Plaza Agreement” and “Semiconductor Agreement”, etc., policy restrictions are imposed on Japan. At the same time, after the rise of the laptop computer industry, the United States focused on cultivating emerging partners, such as Samsung in South Korea and TSMC in Taiwan, and attacked Japanese semiconductor companies through new division of labor.

But in the mid-1990s, Japanese semiconductor companies still had an important role. In 1995, among the top ten global semiconductor companies, NEC ranked first, Toshiba ranked second, Hitachi ranked third, Fujitsu ranked eighth, and Mitsubishi Electric ranked ninth.

However, by 2005, Toshiba had retreated to fourth and NEC to tenth. In 2018, Toshiba, the only top ten player, also had to sell its semiconductor division. As a result, Japanese semiconductor companies have basically withdrawn from the competition in the world market.

In the fourth industrial revolution, the most critical technology is artificial intelligence and other related technologies, such as the internet of Things, blockchain, supercomputing, brain science, etc. However, because the role of artificial intelligence is very significant, this revolution can be called the intelligent revolution. It means that artificial intelligence is not a simple technology, but a strategic technology, which is a key force in a new round of technological revolution and industrial transformation.

What new changes will occur in the fields of transportation, medical care, and education

In real life, we will feel the impact of artificial intelligence more and more intuitively. Here, just a few examples.

Artificial intelligence makes autonomous driving possible. Future transportation will reflect three more elements: first, new energy, second, unmanned driving, and third, shared travel. Due to the advent of driverless technology, people may no longer need to own a car, but directly purchase travel services from origin to destination. As a result, current automakers may need to shift to the role of driverless service operators in the future. In this process, automakers need to fully integrate and cooperate with existing taxi software and high-definition map companies.

Shared mobility can also solve many structural problems. For example, parking difficulty has always been a persistent problem in urban governance, but it is likely to be solved in the context of shared travel. Because at that time, most of the cars will be on the road, so that the parking lot resources can be freed up for greening or places for people to rest.

Medical care is also a problem that plagues human beings. The biggest problem is that medical resources are never enough relative to demand. Research shows that artificial intelligence can greatly expand this resource. In particular, it could gradually automate some tasks traditionally performed by established physicians.

For example, in the field of imaging, artificial intelligence can help radiologists more powerfully. The idea here must be an aid, not a substitute. Many doctors spend a lot of time doing low-end, repetitive tasks. Therefore, when this part of the work is taken over by artificial intelligence, doctors can concentrate more on scientific research and communicate with patients more patiently, so that more targeted treatment plans can be implemented according to the detailed conditions of patients.

In addition, artificial intelligence can also expand medical resources to poorer places. As long as the technology of artificial intelligence achieves real breakthroughs, the solutions it provides will be more stable.

At the same time, based on 5G technology, telemedicine will be more reliable and convenient. Seriously ill patients can be treated by remote devices in isolated areas. We can also use artificial intelligence technology for drug mining, so as to develop new drugs for treating intractable diseases at a lower cost and in a shorter time.

The greatest significance of artificial intelligence in education is that it can promote adaptive education methods. The current education is still “multi-person side”, because the teacher’s time and energy are limited, it is difficult to teach each child according to the personality and characteristics of each child. However, with the help of the artificial intelligence system, we can deeply understand the educational status of each child, and through the adaptive system, we can launch personalized education programs more accurately, so as to help children improve their learning efficiency.

In addition, AI can also play a greater role in teacher assistance, academic assessment, and more.

Will the “gap” between developed and developing countries be closed?

From a global perspective, the development of artificial intelligence may have an important impact on the structure of the future world.

U.S. capabilities and power in AI remain strong. For example, the United States has a very strong advantage in general-purpose computing hardware. Nvidia’s GPU, Google’s TPU, Qualcomm’s smartphone chips, etc. have obvious advantages. For another example, American companies and universities also have a very good tradition in the algorithm framework of artificial intelligence. In some application scenarios, such as the field of special Robots, the technical advantages of Boston Dynamics are also obvious.

In recent years, China has developed rapidly in the field of artificial intelligence. In particular, we are ahead of many countries in the layout of 5G communications. China’s biggest advantage is its huge market and relatively neat consumer groups. China has a large population and a high degree of informatization, which is the basis for the rapid application and implementation of artificial intelligence in various scenarios in the future.

But overall, we still have some deficiencies in basic chips, algorithm frameworks, and ecology. In particular, there is a lot of room for improvement in basic theory and original innovation.

In the new wave of technological revolution and industrial change, some developing countries may be in an awkward and contradictory position. On the one hand, developing countries can use new learning methods and open source software to cultivate relevant talents, greatly shortening the time to catch up with developed countries. At the same time, this intelligent method also helps to overcome the limitations of traditional culture. On the other hand, the intelligent revolution has the potential to further widen the gap between developing and developed countries.

The key problem is that some western developed countries are reluctant to transfer cutting-edge technologies such as artificial intelligence to developing countries; when developing countries achieve breakthroughs in certain emerging technology fields, relevant countries will even use various methods, including Investment screening, export controls, restrictions on technology and personnel exchanges, etc., to hinder the technological progress of emerging countries.

In this sense, the “digital divide” between developed and developing countries may turn into a “smart divide”. Among them, those developing countries that mainly rely on labor as their competitive advantage may be in a more marginal position. Worse yet, as developed countries pull back their industries and drive production through machines, developing countries with poorer foundations will have fewer and fewer opportunities to participate in international competition and world production.

At present, the relevant legal rules, policies and principles of artificial intelligence in the world are mainly defined by Western developed countries. For example, the most influential “Asiloma 23 Principles of Artificial Intelligence” were promoted and formed by Western entrepreneurs such as Musk; Asimov’s “Three Laws of Robotics”, which was influential in the field of artificial intelligence, was also promoted by American science fiction It was first proposed by fiction writers and has become an important principle in the field of machine ethics.

In June this year, China released the “New Generation Artificial Intelligence Governance Principles”. This is the first time a developing country has proposed AI-related governance guidelines, which is of great significance. As China’s strength in the field of artificial intelligence continues to grow, similar rule-making will receive more and more attention and further benefit all mankind.