Tan Tieniu, Academician of the Chinese Academy of Sciences: The past, present and future of AI

Tan Tieniu, Academician of the Chinese Academy of Sciences: The past, present and future of AI

Like the steam engine in the steam age, the generator in the electrical age, and the computer and the internet in the information age, artificial intelligence is becoming the decisive force to push mankind into the age of intelligence. The global industry is fully aware of the significance of artificial intelligence technology leading a new round of Industrial transformation, and has transformed and developed one after another, grabbing the beach and laying out the artificial intelligence innovation ecology. All major developed countries in the world regard the development of artificial intelligence as a major strategy to enhance national competitiveness and maintain national security, and strive to gain dominance in international scientific and technological competition.

Accelerating the development of a new generation of artificial intelligence is a strategic issue that concerns whether my country can seize the opportunity of a new round of technological revolution and industrial transformation. If you miss an opportunity, you may miss an entire era. The dawn of a new round of scientific and technological revolution and industrial transformation is already visible. In this arena that is related to the future and destiny, we must seize opportunities, catch up, and strive to surpass.

Concept and History

To understand where AI is going, we must first know where AI came from.In the summer of 1956, McCarthy, Minsky and other scientists held a meeting at Dartmouth College in the United States to discuss “how to use machines to simulate human intelligence”, and proposed the concept of “Artificial Intelligence (AI)” for the first time, marking the artificial intelligence of artificial intelligence. The birth of intelligence.

Artificial intelligence is a new technical science that researches and develops theories, methods, technologies and application systems that can simulate, extend and expand human intelligence. The research purpose is to promote intelligent machines to listen (speech recognition, machine translation, etc.) Image recognition, text recognition, etc.), speaking (speech synthesis, man-machine dialogue, etc.), thinking (man-machine game, theorem proving, etc.), learning (machine learning, knowledge representation, etc.), acting (Robots, autonomous driving, etc.) car, etc.).

Artificial intelligence is full of unknown exploration road twists and turns. How to describe the development of artificial intelligence since 1956 for more than 60 years, the academic world can be said to be benevolent and wise see wisdom.We divide the development process of artificial intelligence into the following 6 stages:

First, the initial stage of development:1956-early 1960s. After the concept of artificial intelligence was proposed, a number of eye-catching research results have been successively achieved, such as machine theorem proving, checkers programs, etc., setting off the first climax of artificial intelligence development.

The second is the period of reflection and development:1960s-early 1970s. Breakthroughs in the early stages of AI development have greatly raised expectations for AI, and people have begun to try more challenging tasks and put forward some unrealistic R&D goals. However, the succession of failures and the failure of the expected goals (for example, the inability to use a machine to prove that the sum of two continuous functions is a continuous function, machine translation makes a joke, etc.), the development of artificial intelligence has reached a low point.

The third is the application development period:Early 1970s-mid 1980s. The expert system that appeared in the 1970s simulates the knowledge and experience of human experts to solve problems in specific fields, and realizes a major breakthrough in artificial intelligence from theoretical research to practical application, from general reasoning strategy discussion to the use of specialized knowledge. Expert systems have achieved success in the fields of medicine, chemistry, and geology, pushing artificial intelligence into a new climax of application development.

The fourth is the downturn development period:The mid-1980s to the mid-1990s. With the continuous expansion of the application scale of artificial intelligence, the problems of expert systems such as narrow application fields, lack of common sense knowledge, difficulty in acquiring knowledge, single reasoning method, lack of distributed functions, and difficulty in compatibility with existing databases are gradually exposed.

The fifth is the period of steady development:Mid-1990s – 2010. Due to the development of network technology, especially Internet technology, innovative research on artificial intelligence has been accelerated, and artificial intelligence technology has been further promoted to practical use. In 1997, the International Business Machines Corporation (IBM) Deep Blue supercomputer defeated the world chess champion Kasparov. In 2008, IBM proposed the concept of “Smarter Earth”. The above are the landmark events of this period.

Six is ​​the period of vigorous development:Since 2011. With the development of information technologies such as big data, cloud computing, the Internet, and the Internet of Things, computing platforms such as ubiquitous perception data and graphics processors have promoted the rapid development of artificial intelligence technology represented by deep neural networks, which has greatly crossed the gap between science and application. The “technical gap” between the two countries, artificial intelligence technologies such as image classification, speech recognition, knowledge quiz, man-machine game, and unmanned driving have achieved technological breakthroughs from “unusable, unusable” to “usable”, ushering in A new climax of explosive growth.

Status and Impact

There is some “hype” in the society about the development status of artificial intelligence.For example, it is believed that the intelligence level of artificial intelligence systems will soon surpass the level of human beings, that robots will rule the world in 30 years, and humans will become slaves of artificial intelligence, and so on. These intentional or unintentional “hype” and misunderstandings will adversely affect the development of artificial intelligence. Therefore, to formulate strategies, guidelines and policies for the development of artificial intelligence, we must first accurately grasp the current situation of artificial intelligence technology and industrial development.

Dedicated artificial intelligence has made important breakthroughs. From the perspective of applicability, artificial intelligence can be roughly divided into dedicated artificial intelligence and general artificial intelligence. Specialized artificial intelligence systems for specific tasks (such as playing Go) have formed a single-point breakthrough in the field of artificial intelligence due to single tasks, clear requirements, clear application boundaries, rich domain knowledge, and relatively simple modeling. can surpass human intelligence. Recent advances in artificial intelligence have largely been in the area of ​​specialized intelligence. For example, AlphaGo defeated a human champion in the game of Go, artificial intelligence programs surpassed human levels in large-scale image recognition and face recognition, and artificial intelligence systems diagnosed skin cancer to the level of professional doctors.

Artificial General Intelligence is still in its infancy. The human brain is a general-purpose intelligent system, which can draw inferences from one case and integrate it. It can deal with various problems such as vision, hearing, judgment, reasoning, learning, thinking, planning, and design. A truly complete artificial intelligence system should be a general intelligent system. At present, although breakthroughs have been made in the field of dedicated artificial intelligence, the research and application of general artificial intelligence still has a long way to go, and the overall development level of artificial intelligence is still in its infancy. The current artificial intelligence systems have made remarkable progress in “shallow intelligence” such as information perception and machine learning, but their capabilities in “deep intelligence” such as conceptual abstraction and reasoning decision-making are still weak. Generally speaking, the current artificial intelligence system can be described as having intelligence but not intelligence, having IQ but not emotional intelligence, being able to calculate but not “calculating”, and having specialists but not generalists. Therefore, artificial intelligence still has obvious limitations, and there are still many “can’t”, which are far from human intelligence.

Artificial intelligence innovation and entrepreneurship are in full swing. The global industry is fully aware of the significance of artificial intelligence technology leading a new round of industrial transformation, and has adjusted development strategies one after another. For example, Google clearly stated at its 2017 annual developer conference that its development strategy should be shifted from “mobile first” to “artificial intelligence first”, and Microsoft’s 2017 annual report for the first time included artificial intelligence as the company’s development vision. The field of artificial intelligence is at the forefront of innovation and entrepreneurship. The McKinsey & Company report pointed out that in 2016, the global artificial intelligence research and development investment exceeded 30 billion US dollars and was in a stage of rapid growth; the report of CB Insights, a world-renowned venture capital research organization, showed that in 2017, there were 1,100 newly established artificial intelligence start-up companies in the world, and the field of artificial intelligence received a total of The investment was US$15.2 billion, a year-on-year increase of 141%.

The innovative ecological layout has become a strategic highland for the development of the artificial intelligence industry. The development history of information technology and industry is the replacement history of new and old information industry giants rushing to the beach to lay out the innovation ecology of the information industry. For example, the representative companies of the traditional information industry include Microsoft, Intel, IBM, Oracle, etc., and the representative companies of the information industry in the Internet and mobile Internet era include Google, Apple, Facebook, Amazon, Alibaba, Tencent, Baidu, etc. The AI ​​innovation ecosystem includes vertical technology ecosystems such as data platforms, open source algorithms, computing chips, basic software, and graphics processors, and horizontal business and application ecosystems such as smart manufacturing, smart medical care, smart security, smart retail, and smart home. At present, the information industry pattern in the era of intelligent technology has not yet formed a monopoly. Therefore, the global technology industry giants are actively promoting the R&D layout of artificial intelligence technology ecology, and do their best to seize the commanding heights of artificial intelligence-related industries.

The social impact of artificial intelligence is becoming increasingly prominent. On the one hand, artificial intelligence, as the core force of a new round of technological revolution and industrial transformation, is promoting the upgrading of traditional industries, driving the rapid development of the “unmanned economy”, and has a positive impact on people’s livelihoods such as intelligent transportation, smart home, and smart medical care. . On the other hand, issues such as personal information and privacy protection, intellectual property rights of AI-created content, possible discrimination and prejudice in AI systems, traffic regulations for driverless systems, brain-computer interfaces, and technology ethics of human-machine symbiosis have emerged. out, need to hurry to provide solutions.

Trends and Outlook

After more than 60 years of development, artificial intelligence has made important breakthroughs in the “three calculations” of algorithms, computing power (computing power) and calculation materials (data), and is at a technological inflection point from “unusable” to “usable”. But there are still many bottlenecks from “easy to use”.So in the foreseeable future, what trends and characteristics will the development of artificial intelligence show?

From dedicated intelligence to general intelligence. How to realize the leap-forward development from dedicated artificial intelligence to general artificial intelligence is not only an inevitable trend in the development of next-generation artificial intelligence, but also a major challenge in the field of research and application. In October 2016, the National Science and Technology Council of the United States released the National Artificial Intelligence Research and Development Strategic Plan, proposing that the research on general artificial intelligence should be emphasized in the medium and long-term development strategy of artificial intelligence in the United States. Demis Hassabis, the founder of the AlphaGo system development team, proposed to move towards the goal of “creating general artificial intelligence that solves all problems in the world”. Microsoft established the General Artificial Intelligence Laboratory in 2017, and many scientists in perception, learning, reasoning, natural language understanding, etc. participated in it.

From artificial intelligence to human-machine hybrid intelligence. Drawing on the research results of brain science and cognitive science is an important research direction of artificial intelligence. Human-machine hybrid intelligence aims to introduce human functions or cognitive models into artificial intelligence systems, improve the performance of artificial intelligence systems, and make artificial intelligence a natural extension and extension of human intelligence, and solve complex problems more efficiently through human-machine collaboration. . In my country’s new generation of artificial intelligence plan and the American brain plan, human-machine hybrid intelligence is an important research and development direction.

From “artificial + intelligence” to autonomous intelligent system. A large amount of research in the field of artificial intelligence currently focuses on deep learning, but the limitation of deep learning is that it requires a lot of manual intervention, such as artificial design of deep neural network models, manual setting of application scenarios, manual collection and labeling of a large amount of training data, and manual adaptation by users. Intelligent systems, etc., are very time-consuming and labor-intensive. Therefore, researchers have begun to focus on autonomous intelligence methods that reduce manual intervention and improve the autonomous learning ability of machine intelligence for the environment. For example, the follow-up version of the AlphaGo system, Alpha Yuan, starts from scratch and realizes “general chess artificial intelligence” for Go, chess, and Japanese chess through self-play reinforcement learning. In terms of automated design of artificial intelligence systems, the Automated Learning System (AutoML) proposed by Google in 2017 attempts to reduce personnel costs by automatically creating machine learning systems.

Artificial intelligence will accelerate cross-penetration with other disciplines. Artificial intelligence itself is a comprehensive frontier discipline and a highly interdisciplinary compound discipline. Its research scope is extensive and complex. Its development needs to be deeply integrated with disciplines such as computer science, mathematics, cognitive science, neuroscience and social science. With breakthroughs in technologies such as super-resolution optical imaging, optogenetic regulation, transparent brain, and somatic cell cloning, the development of brain and cognitive science has opened up a new era, enabling large-scale and finer analysis of the neural circuit basis and nature of intelligence. Mechanism, artificial intelligence will enter the stage of biologically inspired intelligence, relying on the discovery of biology, brain science, life science and psychology and other disciplines to turn the mechanism into a computable model, and artificial intelligence will also promote brain science, cognition The development of science, life sciences and even traditional sciences such as chemistry, physics, and astronomy.

The AI ​​industry will flourish. With the further maturity of artificial intelligence technology and the increasing investment from the government and industry, the cloudification of artificial intelligence applications will continue to accelerate, and the scale of the global artificial intelligence industry will enter a period of rapid growth in the next 10 years. For example, in September 2016, the consulting firm Accenture released a report stating that the application of artificial intelligence technology will inject new impetus into economic development and increase labor productivity by 40% on the existing basis; by 2035, the United States, Japan, Britain, The average annual economic growth rate of 12 developed countries such as Germany and France can be doubled. A 2018 McKinsey & Company research report predicts that by 2030, about 70% of companies will employ at least one form of artificial intelligence, and the new economic scale of artificial intelligence will reach $13 trillion.

Artificial intelligence will push human beings into an inclusive intelligent society. The innovative model of “artificial intelligence + X” will become more and more mature with the development of technology and industry, have a revolutionary impact on productivity and industrial structure, and promote human beings to enter an inclusive intelligent society. In 2017, the international data company IDC pointed out in the white paper “Information Flow Leads the New Era of Artificial Intelligence” that artificial intelligence will improve the operation efficiency of various industries in the next five years. The transformation and upgrading of my country’s economy and society has a major demand for artificial intelligence. Driven by the demand of consumption scenarios and industrial applications, it is necessary to break the perception bottleneck, interaction bottleneck and decision-making bottleneck of artificial intelligence, and promote the integration and improvement of artificial intelligence technology and all walks of life in society. Build a number of benchmarking innovations in application scenarios to achieve a low-cost, high-efficiency, and wide-ranging inclusive intelligent society.

International competition in the field of artificial intelligence will become increasingly fierce. At present, the international competition in the field of artificial intelligence has kicked off and will become more and more intense. In April 2018, the European Commission planned to invest US$24 billion in artificial intelligence from 2018 to 2020; the French President announced the French Artificial Intelligence Strategy in May 2018, with the aim of welcoming a new era of artificial intelligence development and making France an artificial intelligence. Smart country; In June 2018, Japan’s “Future Investment Strategy 2018” focused on promoting the construction of the Internet of Things and the application of artificial intelligence. The world’s military powers have also gradually formed a competitive situation centered on accelerating the development of intelligent weapons and equipment. For example, the first “National Defense Strategy” report released by the Trump administration of the United States sought to maintain military advantages through technological innovations such as artificial intelligence and ensure that the United States Win future wars; Russia proposed in 2017 that the military industry should embrace “intelligence” to multiply the power of “traditional” weapons such as missiles and drones.

The sociology of AI will be on the agenda. In order to ensure the healthy and sustainable development of artificial intelligence and make its development achievements benefit the people, it is necessary to systematically and comprehensively study the impact of artificial intelligence on human society from the perspective of sociology, formulate and improve artificial intelligence laws and regulations, and avoid possible risks. In September 2017, the United Nations Crime and Justice Research Institute (UNICRI) decided to set up the first United Nations Center for Artificial Intelligence and Robotics in The Hague to regulate the development of artificial intelligence. The White House has organized many seminars and consultations on legal and regulatory issues in the field of artificial intelligence. Industry giants such as Tesla have spearheaded establishments such as OpenAI, which aim to “promote and develop friendly artificial intelligence in a way that benefits all of humanity.”

Situation and thinking

At present, the overall situation of my country’s artificial intelligence development is good. However, we must also be soberly aware that there are risks of overheating and foaming in the development of artificial intelligence in my country, especially in basic research, technical systems, application ecology, innovative talents, legal norms, etc. There are still many problems worthy of attention.In general, the status quo of artificial intelligence development in my country can be summarized as “high attention, gratifying situation, not small gap, and promising prospects”.

highly valued.The Party Central Committee and the State Council attach great importance to and strongly support the development of artificial intelligence. In July 2017, the state released the “New Generation Artificial Intelligence Development Plan”, which placed the new generation of artificial intelligence at the national strategic level for deployment, and described my country’s artificial intelligence development roadmap for 2030, aiming to build a first-mover advantage in artificial intelligence. , grasp the strategic initiative of the new round of scientific and technological revolution. National ministries and commissions such as the National Development and Reform Commission, the Ministry of Industry and Information Technology, the Ministry of Science and Technology, and the Ministry of Education, as well as local governments such as Beijing, Shanghai, Guangdong, Jiangsu, and Zhejiang, have launched policies to encourage the development of artificial intelligence.

Pleasant situation.According to statistics from the “China Artificial Intelligence Development Report 2018” released by Tsinghua University, my country has become the country with the largest investment and financing scale in artificial intelligence in the world. The field of intelligent application is at the forefront of the world. According to the statistics of Elsevier Literature Database in 2017, the number of papers published in the field of artificial intelligence in my country ranks first in the world. In the past two years, the University of Chinese Academy of Sciences, Tsinghua University, Peking University and other colleges and universities have established artificial intelligence colleges. In general, innovation, entrepreneurship, education and scientific research activities in the field of artificial intelligence in my country are very active.

The difference is not small.At present, my country is still in a “follow-up” position in the frontier theoretical innovation of artificial intelligence. Most of the innovations focus on technological applications, and are far from the world leader in basic research, original achievements, top talents, technological ecology, basic platforms, and standards. There are still significant differences in levels. Among the top 700 global AI talents, although China ranks second, it is far lower than the United States, which accounts for about half of the total.In 2018, the market research consultancy Compass Intelligence ranked more than 100 AI computing chip companies in the world, and none of the Chinese companies entered the top ten.In addition, my country’s artificial intelligence open source community and technology ecological layout are relatively lagging behind, the construction of technology platforms needs to be strengthened, and the international influence needs to be improved. my country’s participation in the formulation of international AI standards is not enough, and the formulation and implementation of domestic standards is lagging behind. my country still lacks in-depth analysis of the possible social impact of artificial intelligence, and the process of formulating and improving artificial intelligence-related laws and regulations needs to be accelerated.

The outlook is good.The development of artificial intelligence in my country has comprehensive advantages in terms of market scale, application scenarios, data resources, human resources, smartphone popularization, capital investment, and national policy support, and the development prospects of artificial intelligence are promising. According to the “Artificial Intelligence: Helping China’s Economic Growth” report released by Accenture, the world’s top management consulting firm, in 2017, artificial intelligence is expected to boost China’s labor productivity by 27% by 2035. The “New Generation Artificial Intelligence Development Plan” released by my country proposes that by 2030, the scale of the core artificial intelligence industry will exceed 1 trillion yuan, and the scale of related industries will exceed 10 trillion yuan. In my country’s future development journey, the “smart dividend” will hopefully make up for the lack of demographic dividend.

At present, it is a major historical opportunity period for my country to strengthen the layout of artificial intelligence, reap the dividends of artificial intelligence, and lead the era of intelligence. How to choose China’s path, seize China’s opportunities, and show China’s wisdom in the wave of artificial intelligence booming requires in-depth thinking.

Establish a rational and pragmatic development concept. The development of anything cannot be at a high level all the time. There are highs and lows. This is an objective law. The realization of autonomous intelligence and general intelligence of machines in any real environment still requires medium and long-term theoretical and technical accumulation, and the penetration and integration of artificial intelligence in traditional fields such as industry, transportation, and medical care is a long-term process, which is difficult to achieve overnight. Therefore, the development of artificial intelligence should fully consider the limitations of artificial intelligence technology, fully recognize the long-term and difficult nature of artificial intelligence reshaping traditional industries, rationally analyze the development needs of artificial intelligence, rationally set artificial intelligence development goals, and rationally choose artificial intelligence. Intelligent development path, and pragmatically promote artificial intelligence development measures. Only in this way can we ensure the healthy and sustainable development of artificial intelligence.

Pay attention to the original research that strengthens the foundation and strengthens the foundation.The frontier basic theory of artificial intelligence is the cornerstone of artificial intelligence technology breakthrough, industry innovation, and industrialization promotion. Facing the critical point of development, in order to gain the ultimate right to speak, major breakthroughs must be made in the basic theory and cutting-edge technology of artificial intelligence. In accordance with the requirements of supporting scientists to brave the “no man’s land” at the forefront of artificial intelligence technology, we must strive to achieve transformative and subversive breakthroughs in the development direction of artificial intelligence, theories, methods, tools, and systems, and form artificial intelligence with international influence. The intelligent original theoretical system provides theoretical support for the construction of an independent and controllable artificial intelligence technology innovation ecology in my country.

Build an independent and controllable innovation ecosystem.my country’s artificial intelligence open source community and technological innovation ecological layout are relatively lagging behind, and the construction of technology platforms needs to be strengthened. We must be problem-oriented, focus on key core technologies, accelerate the establishment of a new generation of artificial intelligence key common technology systems, comprehensively enhance artificial intelligence scientific and technological innovation capabilities, and ensure that key core artificial intelligence technologies are firmly in our own hands. We must focus on preventing the risk of “hollowing” in the era of artificial intelligence, systematically deploy and focus on the development of “new core and high base” in the field of artificial intelligence: “new” refers to a new type of open innovation ecology, such as the integration of production, education and research; “nuclear” refers to core key technologies and devices, such as advanced machine learning technology, robust pattern recognition technology, low-power intelligent computing chips, etc.; “High” refers to high-end integrated application systems and platforms, such as machine learning software and hardware platforms, large-scale data platforms, etc.; “Base” refers to Basic theories and methods with great original significance and technological drive, such as brain-computer interface, brain-like intelligence, etc. At the same time, we must pay attention to the construction of artificial intelligence technical standards, and the testing of product performance and system security. In particular, my country is at the forefront of the world in the application of artificial intelligence technology. It should have the right to speak in the formulation of international artificial intelligence standards, and accelerate the process of artificial intelligence-driven economic and social transformation and upgrading through the implementation of standards.

Promote shared global governance.At present, developed countries have controlled the upstream resources of the industrial chain through artificial intelligence technology innovation, and the insurmountable technological gap and industrial barriers may further widen the gap in productivity development between developed and developing countries. Among developing countries, my country is expected to become the leader in the global artificial intelligence competition. It should build an open, shared, high-quality and low-cost artificial intelligence technology and application platform that benefits the world. “Promoting the joint construction of a community with a shared future for mankind.

The Links:   3BSE032401R1 3BSC610042R1/SS822

Pre:    Next: