In-depth analysis of the artificial intelligence development strategies of these 10 countries

At present, almost all major countries in the world have formulated their own AI strategies.

This article will sort out the strategic thinking of more than a dozen countries in the world that are in a relatively leading position and have clearly proposed national AI development strategies, including China, the United States, France, Japan, Singapore, etc. This will help us estimate the development prospects of AI .

middle  country

China may become an AI powerhouse by 2030 , but is currently still catching up with the United States in terms of total investment, talent resources, and experience in the field of AI . In 2012-2016 :

1. China ’s total investment in AI is $ 2.6 billion, while the United States’ is $ 17.2 billion.

2. China has 39,000 AI talent resources , while the United States has 78,000 .

In addition, as the world's most populous country, China has a wealth of information and data. It is estimated that by 2020 , China will have 20% of the world's data (equivalent to 44 ZB ). To this end, the government has announced the "New Generation AI Development Plan", which contains the following details:

Promote close cooperation between the government, academia and industry to formulate strategic goals for AI development and achieve progress:

• 2020 : AI overall technology and applications will be in sync with the world's advanced level. The scale of AI core industry will exceed US$ 22.5 billion, driving the scale of related industries to exceed US$ 150.8 billion.
• 2025 : AI basic theory achieves major breakthroughs, and some technologies and applications reach world-leading levels. The scale of the AI ​​core industry exceeds US$ 60.3 billion, driving the scale of related industries to exceed US$ 754 billion.
• 2030 : AI theory, technology and applications will reach the world's leading level, and China will become the world's major AI innovation center. The scale of the AI ​​core industry will exceed US$ 150.8 billion, driving the scale of related industries to exceed US$ 1.5 trillion.

Identify and fully implement six key tasks, including:

• Build an open and collaborative AI technology innovation system: establish a new generation of AI basic theoretical system; establish a new generation of AI key common technology system; coordinate the layout of AI innovation platforms; and accelerate the cultivation and gathering of high-end AI talents.
• Cultivate a high-end and efficient intelligent economy: vigorously develop emerging AI industries; accelerate the intelligent upgrading of industries; vigorously develop smart enterprises; and build an AI innovation highland.
• Build a safe and convenient intelligent society: develop convenient and efficient intelligent services; promote intelligent social management; use AI to enhance public security protection capabilities; and promote social interaction, sharing and mutual trust.
• Strengthen military-civilian integration in the field of AI .
• Build a ubiquitous, safe and efficient intelligent infrastructure system.
• Make forward-looking plans for major new-generation AI scientific and technological projects.

Nine AI technology fields have been identified, including one AI full technology field and eight AI technology fields :

AI technology foundation areas:

• Basic research on deep learning, neuroinformatics, neural information processing systems, etc.
• Research and application in computer vision, biometrics, complex environment recognition, human-computer interaction, natural language processing, automatic translation, intelligent control and network security.

Eight AI technology areas:

• Public Service Platform Computing
• Smart Home
• unmanned
• Intelligent transportation applications
• Smart Security
• AI terminal applications
• Wearable devices
• robot

In addition, four national drivers for the development of AI were identified , including hardware, research, algorithms, and the AI ​​business ecosystem:

1. Hardware: Advocate to catch up with advanced countries in chip and supercomputer manufacturing. Promote competition with Chinese methods and combinations, encourage transactions with foreign companies, and encourage technology giants and startups to build supercomputers and invest in the production of AI chips.

2. Data: Emphasize the advantages of promoting data sharing between governments and businesses when acquiring data, and promote the protection of data flow at different levels by regulating AI -related industries and strengthening discussions against the abuse of data at the commercial level as people increasingly pay attention to the privacy risks brought by AI .

3. Develop algorithms: On the one hand, attract and cultivate talents (especially the world's top AI talents) by supporting basic research; on the other hand, encourage technology companies such as Baidu, Huawei, Alibaba, Tencent or iFLYTEK to establish AI research institutes overseas to recruit AI talents to overcome the problem of low paper results and AI education quality.

4. Build an AI business ecosystem: Invest more than $ 1 billion in domestic start-ups and guide local governments and state-owned enterprises to attract private investment to fund AI projects to obtain data from society and align corporate goals with national development plans.

beautiful  country

The U.S. government believes that AI is a promising transformational technology with tremendous economic and social benefits. AI can revolutionize how Americans live, work, learn, research, and communicate. In addition, AI research can promote economic prosperity, increase educational opportunities and quality of life, and improve national security. Because of these potential benefits, the U.S. government has been investing in AI research for many years.

On May 3, 2016 , the U.S. government announced the creation of a new NSTC subcommittee on machine learning and AI to help coordinate federal efforts in AI . On June 15 , 2016 , a subcommittee of the Network and Information Technology Research and Development Program ( NITRD) was appointed to plan a national AI R&D strategy. They then established the NITRD AI Group to determine strategic priorities for federal AI R&D, with a special focus on areas that businesses cannot handle.

The National AI R&D Strategic Plan sets a series of goals for federally funded AI research, including both internal government research and federally funded research outside the government, such as at research institutes, universities, etc. This research aims to create new AI knowledge and technologies that provide many benefits to society while minimizing negative impacts. To achieve this goal, the AI ​​R&D Strategic Plan outlines the priorities for federally funded AI research:

Strategy 1 : Make long-term investments in AI research. Prioritizing investments in the next generation of AI will lead to greater discovery and understanding, allowing the United States to continue to lead the world in AI .

Strategy 2 : Develop effective human-machine collaboration methods. AI systems will not replace humans, but will work with humans to achieve optimal performance. Therefore, it is necessary to study how to establish effective interactions between humans and AI systems.

Strategy 3 : Understand and address the ethical, legal, and social aspects of AI . AI technology needs to work within the context of formal and informal standards. Therefore, research is needed to understand the ethical, legal, and social impacts of AI . At the same time, AI systems should be developed and designed to meet ethical, legal, and social goals .

Strategy 4 : Ensure the safety of AI systems. Before AI systems are widely used, it is necessary to ensure that they operate safely in a controllable, clear, and understandable manner. This requires further research to address this challenge by building reliable and trustworthy AI systems.

Strategy 5 : Develop sharable public datasets and environments for AI training and testing. The depth, quality, and accuracy of datasets and training resources affect the performance of AI . Therefore, researchers should develop high-quality datasets and environments and provide responsible access to high-quality datasets and testing and training resources.

Strategy 6 : Measure and evaluate AI technologies through standards and benchmarks . AI progress, levels, standards, demonstrations, and community participation are key to guiding and measuring AI development. More research is needed to develop various evaluation techniques.

Strategy 7 : Better understand the nation’s AI R&D workforce needs. Advances in AI require a strong AI research community. Therefore, increasing knowledge about the current and future AI R&D workforce will help ensure there are enough AI experts to address the strategic R&D areas outlined in this plan.

The entire federal government can accomplish the plan’s seven strategies and realize its vision by supporting the following recommendations:

Recommendation 1 : Building on Strategies 1-6 , the development of an AI R&D framework should take into account potential market opportunities and more effectively integrate collaboration between AI R&D and investment: Federal agencies should work together through NITRD to develop a R&D framework to promote the coordination and progress of the R&D efforts mentioned in this plan. This will enable agencies to plan, coordinate and collaborate effectively to support this strategic plan. The implementation framework should take into account the R&D priorities of each agency based on its mission, capabilities, authority, and budget. Based on this implementation framework, it may be necessary to establish some funding programs to accelerate the AI ​​R&D agenda. To help implement this strategic plan, NITRD should consider forming an inter-agency working group specifically to promote collaboration among different agencies.

Recommendation 2 : In line with Strategy 7 of this plan , study the feasibility of creating and maintaining a healthy AI research and development workforce at the national level: A healthy and vibrant AI R&D workforce is critical to addressing the strategic challenges of research and development outlined in this report. While some reports indicate that there may be a shortage of experts in AI research, there are no official workforce data reports that describe the current state of the AI ​​workforce, projected workforce entry plans, or the comparison of AI workforce supply and demand forces. NITRD should study how to describe and define current and future AI R&D workforce needs as accurately as possible and develop additional research or recommendations to ensure that there is an adequate R&D workforce to meet the nation’s AI needs. Recommendation based on the findings: Federal agencies should ensure that they can establish and maintain a healthy national AI research and development workforce.

Canada

Canada was one of the first countries to release an AI strategy. In the 2017 federal budget, a five-year plan called the Pan-Canadian AI Strategy was released , which included $ 125 million for investments in research and AI . The strategy has four goals :

1. Increase the number of scientists and graduates;

2. Identify three groups of outstanding scientists;

3. Develop thought leadership on the economic, ethical, political, and legal implications of AI .

4. Support the national AI research community.

The Canadian Institute for Advanced Research led the strategy and is working closely with the Canadian government and three new AI institutes: the Alberta Machine Intelligence Institute (AMII) in Edmonton , the Vector Institute in Toronto, and the Algorithmic Learning Institute in Montreal.

Law  country

Europe lags behind the United States and China in developing AI development strategies. While Germany is focused on the Fourth Industrial Revolution and the United Kingdom is focused on Brexit, French President Emmanuel Macron announced that the government has endorsed a national “AI leadership ” strategy and will invest 1.5 billion euros over 5 years ( 2018-2022 ) as a representative of the European national AI strategy. The French President’s statement on AI development strategy summarizes the key points of the French and European AI strategy report prepared by Cédric Villani (French mathematician, 2010 Fields Medal winner, and French Member of Parliament) and his partners . The seven key elements of the report include:

First, formulate appropriate data policies to encourage enterprises to create and share data, develop data of social interest, and support the right to data backup.

Second, the four major strategic focus areas for AI development are health, transportation, environment, and national defense and security. In response to key issues, policies are formulated in each strategic area to lay the foundation for platforms in specific regions and examine innovation zones in each region.

Third, leverage France’s potential in AI research and development, establish interdisciplinary AI organizations in selected universities and research institutions, allocate appropriate research resources (including supercomputers designed specifically for AI applications in collaboration with manufacturers); increase researchers’ salaries and strengthen exchanges between industry, academia and research.

Fourth, the plan addresses the impact of AI technology on workers, establishes public laboratories to respond to job changes, conducts research on the complementarity of machines and humans, and evaluates new methods of vocational training.

Sixth, ensure the transparency of AI development technology, establish a clarity and algorithm review system, pay attention to the responsibility of AI agents for ethical threats, establish private ethics committees related to digital and AI technologies, organize public debates on AI ethics, and adhere to the principle of human obligations (mainly when using AI tools in public services ).

Australia

Australia does not yet have a clear AI strategy. However, in Australia 's 2018-2019 budget , the government announced an investment of A$ 29.9 million over four years to support the development of AI . In addition, the government has developed a technology roadmap, a standards framework, and a national AI ethics framework to support the development of the issue responsibly. Non-governmental organizations also support joint research center projects, doctoral scholarships, and other initiatives to improve Australia's AI talent supply.

Furthermore, in the 2017 innovation roadmap, Australia 2030 : Prosperity through Innovation, “ the government announced that it would prioritise AI in the government’s digital economy strategy , which was released in the second quarter of 2018. ”

Germany  country

Germany's AI plan calls for an investment of 3 billion euros by 2025 , and hopes to double that amount by partnering with private companies to achieve the goal of making " Germany and Europe an AI hub . " Germany's national AI plan focuses on developing AI centers across the country , investing in education and attracting the next generation of AI talent to prepare for data digitization.

Germany has set a number of output targets, such as the formation of a national network of 12 new AI research centers, 100 new AI teaching positions, and a government funding program: providing AI- related support services to 1,000 small and medium-sized enterprises each year. In addition, in order to make broad and sustainable changes to education, the German AI Association has made some policy recommendations, including the introduction of compulsory education in data science in the third year. In addition, the government will decide on internships and academic courses at universities. Therefore, when the government released the National AI Plan in 2018 , the German states rushed to set up new organizations and prepare to apply for federal funds.

The government has also called for attracting German researchers working abroad. Germany plans to recruit 30 new international lecturers ( six new lecturers per year) from 2018 to 2024 through the Alexander von Humboldt Fellowship program. The program provides a start-up investment fund of 3.5 million euros or 5 million euros, depending on the content of the research.

Subsequently, the government also called for a push for digitalization of corporate data. They implemented an integrated “ digitalization center ” through a new consulting program outlined in the AI ​​strategy to support the digitalization of 1,000 SMEs each year.

United Arab Emirates

The UAE government launched its AI strategy in October 2017 , becoming the first country in the Middle East to develop an AI strategy and the first to establish an AI department. The strategy is the first initiative of the UAE Centennial Plan 2071. The main goal of the master plan is to use AI to improve government efficiency.

The government will invest in AI technology in nine sectors : transportation, health, space, renewable energy, water, technology, education and the environment. The government aims to reduce costs across government, diversify the economy and position the UAE as a global leader in AI applications.

day  Book

Japan has always been the world's fourth largest economy by GDP . However, Japan's AI development market has increased from about 3.7 trillion ( 2015 ) to about 87 trillion ( 2030 ). The AI ​​development strategy that leads the world in the following aspects :

First, the government established the Japan AI Technology Development Strategy Committee to vertically manage five national research and development institutions and three core development centers (National Institute of Information Technology, National Center, RIKEN, and National Institute of Industrial Science and Technology).

Secondly, the AI ​​industrialization process focuses on three priority areas, including productivity, healthcare and services. Among them, healthcare is divided into three stages: the first stage ( 2020 ): promote the application of AI direct data to promote the application in related fields; the second stage ( 2020-2025 ): expand the public application of AI and its data to a wider range of fields; the third stage ( 2025-2030 ): establish an AI ecosystem based on the connection and mixing of multiple fields .

Third, the three core R&D centers focus on social AI technologies based on diverse data . Diverse data includes: individuals, voice conversations, internal medicine, action and search history, living and working space, sales and manufacturing, transportation, nature, weather and maps (land, urban areas); AI technologies include: image recognition, natural language processing, speech recognition / synthesis and prediction. Over the past 10 years, the government has tripled the investment in AI R&D by university-affiliated companies and R&D institutions , while also promoting more outstanding private R&D investment.

Finally, create a development environment for young researchers, especially in the first stage to attract high-level AI development talents from home and abroad, and encourage AI researchers to actively participate in the development of AI technology.

Korea  country

In May 2018 , the South Korean Fourth Industrial Revolution Committee announced a national AI development strategy, investing 2.2 trillion won, attracting 5,000 experts, and becoming one of the four major powers in global AI development. The strategy lasts until 2030 and consists of four phases:

Phase 1 ( 2020 )

Core technology: Development of audio-visual comprehension technology. Extended technology: AI question-answering system in professional fields . Reduce the time for the health sector to find new drugs from five years to one year. Background technology: Complex information analysis involves the use of high-power descriptive operations. Attract and cultivate 590 AI senior talents and 2,250 AI ordinary employees . Build 66.7 million shared data, 4.3 million industrial data, and 9.2 billion Korean understandings. Provide supercomputing support for 300 organizations every year.

Phase 2 ( until 2022 )

Basic technologies: Master unsupervised learning theory, image synthesis technology, tracking - detection and prediction technology, and descriptive function inference learning (until 2025 ). Scalable technologies: Real-time risk detection system. Reduce the development cycle of new drugs in the medical industry by more than half (from 15 years to 7 years). Key technologies: Cognitive information exchange between brain neural networks and AI neural networks; overall brain and machine safety (under development, to 2025 ). Attract and train 1,370 senior AI talents and 3,600 AI ordinary employees , build 111 million shared data, 48.5 million industrial data, and 15.3 billion Korean understandings. Provide supercomputing support for 400 organizations every year .

Phase 3 ( until 2025 )

Core technologies: Continue to learn descriptive inference on functions. Commercialize artificial neural network chips. Expanded technologies: Question-answering systems for images. Develop new drugs tailored to each individual. Key technologies: Cognitive information exchange between brain neural networks and AI neural networks; integrated interface for brain and machine safety. Cultivate talents with world-class AI leadership (until 2030 ). Strengthen infrastructure research in the form of improved collaboration (until 2030 ).

Phase 4 ( until 2030 )

Core technology: Autonomous collaboration between AI and humans using unsupervised learning techniques. Extended technology: Providing food and drug preparations tailored to each specific audience. Background technology: Strengthening and improving human cognitive abilities through the application of AI . Cultivating talents with world-class AI leadership. Strengthening research infrastructure through enhanced collaboration.

The choice of investment is to focus on new technologies; it is difficult for the public sector to attract private investment and to establish original markets in areas where private competitiveness is high. Therefore, the motto of practice is to ensure technological capabilities, develop AI according to international standards with low technology in basic science (new generation AI based on cognitive science, neural network computing) , AI chip layer, high-performance AI computing, and application areas according to the AIX formula (new drugs, future materials, industrial applications).

Establish training institutions for AI development graduate students and postdoctoral students, and strengthen AI training and research support in colleges and research institutions . Build public and private AI brain labs, AI centers, and AI infrastructure platforms. Saltlux, South Korea's first AI company, received 32 billion won in investment in AI products.

Singapore

Singapore’s National AI Plan invests $ 150 million over five years to combine national AI capabilities to drive Singapore’s digital economy. The plan has three goals:

1. Use AI to solve key social and industrial problems in areas such as transportation and healthcare.

2. Invest in improving AI capabilities ( AI systems that explain the next generation of AI , cognitive science, AI talent training, etc.)

3. Provide 100 projects to promote the application of AI and machine learning in industry.

The strengths and weaknesses of national AI strategies around the world

Generally speaking, AI strategies announced at the national level focus on AI training and talent attraction. However, no country has yet provided more detailed guidance on career guidance for AI training.

Most of the recommendations come from workshops at educational institutions and private companies, so there is no consensus on this issue. At present, some of the advantages and disadvantages of the National AI Strategy 2020-2030 in terms of AI career guidance are as follows.

Advantages:

1. Focus on key economic entities and sectors to promote the implementation of the national AI strategy.

2. Help determine the scale of impact that AI technology will have on each country’s economy.

3. Help identify the challenges and benefits of AI , how to maximize competitive advantage, and provide solutions to overcome the obstacles that AI poses to the economy.

4. Clearly define the development roadmap for achieving established strategic goals.

Disadvantages:

1. All predictions about the benefits of AI are still somewhat vague, subjective, and have no clear scientific basis.

2. National AI strategies are very general and challenging to implement.

3. The shortage of talent resources in AI remains a challenge for education planners in every country.

Strategic priorities for AI development

What is unique about policy development in this area is that the approaches taken by governments around the world to promote the use and development of AI are quite wide-ranging. Not only are they advancing different policies, but they are also focusing on different areas of public policy.

This framework provides a rough classification of AI strategies in public policy areas and assesses the relationship between AI strategic priorities and their research funding and attention via a heat map (see explanation below).

The development of a global inventory of AI strategies is hampered by two challenges. First, different AI strategies vary widely; they may come from a website, an official white paper, a working report, or a budget announcement. Therefore, it is possible that a strategy may be overlooked due to the rapid and diverse developments in the field.

Second, some governments announced new initiatives after releasing their original approaches. To allow for a more systematic analysis of each plan, this analysis focuses only on those that were first announced.

Finally, the policy announcements for each strategy were categorized into eight public policy areas:

1. Scientific research: Establish new research centers or programs for basic and applied AI research, or commit to increasing existing public research funding for AI .

2. AI Talent Development: Provide funding to attract, retain, and train AI talent at home and abroad, including funding certain leading talents or establishing specialized AI master's and doctoral programs.

3. Technology and job prospects: Helping students and the entire labor market develop relevant job skills, such as investing in STEM (science, technology, engineering and mathematics) education, digital skills, or lifelong learning.

4. Industrialization of AI technologies: Encourage the private sector to adopt AI technologies, including investment in strategic sectors, funding for AI start-ups and SMEs, and strategies to create AI clusters or ecosystems.

5. Ethical standards for AI : Establish a council, committee, or working group to develop standards or regulations for the ethical use and development of AI . This area also includes specific funding for research or pilot projects to develop explainable and transparent AI .

6. Data and digital infrastructure: Funding for open data partnerships, platforms, and datasets, and committing to creating testing environments and regulatory sandboxes.

7. AI in Government Governance : Establish pilot projects to use AI to improve government efficiency, service delivery, and public management.

8. Inclusion and social well-being: Ensure that AI promotes social inclusion, as well as the inclusion of AI’s own industry background and perspectives.