Opening Address By Mr Heng Swee Keat, Minister For Finance, Singapore, At The Sino-Singapore Artificial Intelligence (Ai) Forum, 1 June 2018, Friday, 9.20am At The Singapore Nanjing Eco Hi-Tech Island Tech Innovation Centre, Nanjing, China01 Jun 2018
a. 这是我过去一年里，第二次访问江苏, 第三次涉足中国。
Impact of Artificial Intelligence (AI)
7. New technologies are rapidly changing the way we live, work and play.
8. One of the most transformative technologies is AI, which can be applied across the economy and society.
a. AI systems can quickly analyse and synthesise huge amount of information. In many areas, AI can already do as well, and sometimes even better than humans. For example,
i. In law enforcement, AI and facial recognition technology can help to catch suspects. I read about the three suspects caught at three different Jacky Cheung concerts in China recently.
ii. In healthcare, experiments have shown that AI can interpret radiographs faster and with an accuracy comparable to radiographers.
iii. In speech recognition and translation, you see a very good example of what’s happening out there. The AI system is able to translate what I’m saying. Perhaps one day, there will be widespread use of AI to translate speeches like this in real time, and enhance human understanding of different languages. I understand that Chinese companies are working hard on this.
iv. AI not only does not deteriorate like traditional capital assets, but in fact, gets better over time.
b. AI can also raise productivity.
i. For countries like Singapore and China, this is especially important given our ageing population and slowing labour force growth.
ii. By taking on the more tedious and mechanical work, AI can improve efficiency while enhancing customer experience.
1) For instance, Ping An Insurance’s Smart Verification AI platform can quickly and accurately verify the identity of insurance applicants or agents. This can reduce the number of disputes, and cut down time required to settle claims from three days to just 30 minutes.
c. AI can also help to improve the lives of our people.
i. I visited 滴滴出行during my last visit to China. I was impressed by how they have leveraged big data and AI to better match passengers with drivers, which has significantly cut waiting times for customers.
9. The economic potential for AI is significant.
a. A recent study by McKinsey estimated that AI can create 22 to 37 trillion RMB (US $3.5 to $5.8 trillion) in annual value for the global economy.
10. AI’s impact is even greater when combined with other technologies.
a. For example, advancements in AI and genomics are transforming the field of medicine. AI helps us interpret large amounts of genomic data to develop genetic medicines for complex diseases.
b. In the manufacturing sector, the use of different technologies together is transforming entire systems of production, management, and governance. Many call this the fourth industrial revolution.
Working Together – Governments, Businesses, Academia and Public
11. The potential is promising, but to realise the potential, governments, businesses, academia and the public have to work together.
a. Academia and businesses can generate new research, ideas, products and services.
b. The government plays the role of an enabler, putting in place the supporting ecosystem to spur innovation.
c. The public must have the skills, trust and confidence to use AI.
12. We need to work together on these three key enablers:
a. First, investing in research and development.
b. Second, creating a robust regulatory environment.
c. Third, promoting acceptance and adoption.
Investing in Research and Development (R&D)
13. Let me start with research and development.
a. To develop good AI solutions, we need to have a deep understanding of the problems that need to be solved, as well as have the expertise and creativity to create new solutions to these problems.
i. Singapore has been investing heavily in R&D.
1) The National Research Foundation (NRF) that I oversee has sustained public R&D spending at 1% of GDP annually. We have seen some good results.
2) Today, based on field-weighted citation impact, Singapore is ranked first in the world for research quality in AI.
b. Outside of academic journals and publications, many brilliant academics are also working with companies to commercialise the use of AI. There are many opportunities for researchers and businesses from both Singapore and China to do more together.
i. For research, we have the Singapore-China Joint Research Programme (JRP).
1) Established since 1999, it provides joint funding for collaborative projects.
2) In 2015, to deepen our collaboration, we had another joint grant call programme between Singapore’s NRF and the National Natural Science Foundation of China (NSFC).
ii. For commercialisation, we have joint projects that marry the research expertise of our academia with the commercial experience of businesses and digital giants, to create market-ready AI solutions.
1) For example, the joint research centre started in 2012, by Singapore’s Institute for Infocomm Research and Baidu, focusing on natural language processing for South East Asian languages.
2) More recently, Singapore’s Nanyang Technological University (NTU) and Alibaba have set up a joint research institute to explore AI solutions in areas such as healthcare, ageing and lifestyle.
3) In fact, the Singapore University of Technology and Design (SUTD) will be signing an MOU today, with iFLYTEK, the Singapore-Nanjing Eco Hi-Tech Island (SNEI) Development Company and the SNEI Administrative Committee. They will explore setting up a joint research lab here to focus on AI application in smart building design and architecture.
iii. We also have the Global Innovation Alliance (GIA).
1) This is a network of Singapore and overseas partners in major innovation hubs and key markets. Our entrepreneurs and businesses can tap on this network to source for innovation, investors, test-bed products, and implement solutions.
2) GIA (Beijing) and BLOCK71 Suzhou, which I opened last year, are two of the latest additions to the GIA network.
3) Today’s forum is an excellent example of how we can bring together people from different backgrounds to generate fresh insights and co-create solutions through new partnerships. I’m happy to see so many participants from Singapore and China!
c. Singapore and China can also collaborate in third party markets.
i. This is aligned with China’s “going out” strategy (走出去战略), to encourage Chinese companies to venture overseas.
ii. Singapore is the ASEAN Chairman and country coordinator for ASEAN-China Dialogue Relations. We have also designated 2018 as the ASEAN-China Year of Innovation.
1) One of our key initiatives is the ASEAN Smart Cities Network. 26 pilot cities in ASEAN are developing action plans to catalyse bankable projects for smart and sustainable urban development. China has about 500 smart cities that have initiated pilot projects. In fact, it is the country with the largest number of smart cities.
2) There is big potential for Singapore and China companies to form partnerships to test-bed innovative AI solutions for the region.
Creating a Robust Regulatory Environment
14. I've spoken on the first enabler of investing in research and development. The second enabler is in creating a robust regulatory environment to promote innovation and adoption.
a. Many see this as counter-intuitive, as rules and regulations may be more traditionally seen to stifle, rather than enable innovation.
b. But my experience running the Monetary Authority of Singapore (our central bank, and integrated financial regulator), I see good regulation as a necessary part of development, and in fact a key enabler.
i. Good regulatory policies build public trust. For technologies to receive widespread acceptance, the public must be convinced that their safety, privacy, and data are safeguarded.
1) Ensuring the security and responsible use of personal data is a key priority, and there is significant public interest in this issue.
2) The EU just implemented the General Data Protection Regulation (GDPR) in May. There is also a lot of interest in its ePrivacy legislation, which is being discussed in the EU. And I understand that there are some more regulations that the EU is looking at.
Singapore has put in place the Personal Data Protection Act since 2014. We have also introduced a new Public Sector (Governance) Act to strengthen safeguards for data sharing among government agencies. Different countries place different values on public security, privacy protection and efficiency of private transactions. And there are many different sets of trade-offs that different societies will have to decide on.
3) Another priority is the ethics of AI. The machine or software has no morals. But many of the decisions it makes have moral dimensions. If an autonomous vehicle, in an emergency, swerves left, it would hit a young child. If it swerves right, it would hit an elderly person. Which way do you programme it to swerve? The decision depends on the algorithm we write into it. The development of AI ethics must keep pace with innovation to build the public trust required for widespread acceptance. This is an area we must continue to discuss.
ii. Regulations also help to level the playing field – for new entrants and start-ups with innovative ideas to compete more fairly with incumbents.
c. But how do we design a robust regulatory framework that promotes experimentation and innovation, while safeguarding the public interest as well as taking the ethical considerations into account?
i. We need to have a keen understanding of emerging technologies and new business models, and to be alert to potential risks.
ii. In Singapore, we do this via regulatory sandboxes. For the duration of the sandbox, legal and regulatory requirements are relaxed, for firms to testbed their new solutions in real but contained environments.
iii. For example, the Monetary Authority of Singapore has a Fintech sandbox for companies to develop new financial solutions before launching them in the market.
iv. Similarly, the Land Transport Authority has an autonomous vehicle (AV) sandbox to prepare for the future deployment of AVs in Singapore.
v. We welcome businesses and entrepreneurs from China to testbed your solutions in our regulatory sandboxes.
d. Given the large amount of investments into AI, there have also been concerns on a possible bubble. We need to be vigilant. While we seek to maximise the potential of AI, we should also seek to develop solutions to address real needs in our society.
Promoting Acceptance and Adoption
15. I've spoken on the first two enablers – investing in R&D, and creating a robust regulatory environment. The third enabler is to promote the widespread acceptance and adoption of AI solutions by the public.
a. To boost adoption of AI among Singapore companies, we invested RMB 700 million (S$150 million) in AI Singapore (AISG) last year.
i. It brings together academia and companies to translate research outcomes into real-world solutions in areas such as healthcare, urban development and finance.
ii. It has engaged over 150 companies over just the past six months. I’m very encouraged by the strong industry interest.
b. Beyond businesses, we must equip our workers and citizens with the relevant skills so that they can adopt AI at work and in their daily lives.
i. For workers, as we re-design jobs, we must help them reskill. In this way, they can take on higher value-add roles and grow their wages.
1) We started the SkillsFuture movement in Singapore – to encourage our people to undertake lifelong learning.
2) Specifically, our TechSkills Accelerator or TeSA helps workers to learn and use emerging technologies such as AI.
ii. We must also help our young people to build a strong foundation of knowledge and skills so that they are ready for the AI-enabled economy of the future. In so doing, we can also identify and groom AI talents from young.
1) Both Singapore and China are taking active steps.
2) I understand that the Chinese Government has recently announced the release of an AI textbook for high school students, with 40 high schools across China participating in the pilot.
3) In Singapore, our secondary schools have introduced computer studies as an “O” level examination subject since 2006. “O” level examination is typically taken by 16-year old students in Singapore.
Our schools also organise enrichment programmes and co-curriculum activities to expose our students to areas such as robotics.
4) At the tertiary level, NTU is launching a new degree programme in data science and AI.
iii. For the wider population, we must ensure digital literacy, so that everyone can benefit from the convenience enabled by AI and other technologies.
1) In Singapore, we provide basic infocomm training in the community for those who are less digitally savvy, so that no one is left behind.
16. So to summarise, the potential for AI is significant. It has the ability to make an impact on our society and our economy, and we will need to work together – academia, businesses, governments and the public, to do three things:
a. Invest in research and development;
b. Create a robust regulatory environment; and
c. Promote acceptance and adoption of AI.
17. So let me now summarise this in Mandarin.
b. 第二,我们创造一个健全的监管环境。监管是成功发展科技的要素之一。它甚至可以推动创新。良好的监管政策和措施也能够建立公信。在新加坡，我们设立了监管沙盒。执行监管沙盒时，相关的法律和监管条例会适当 放松。企业可以试验新的解决方案。我们欢迎中国企业借助我们的监管沙盒来试验新解决方案。
为了鼓励 新加坡企业采纳 人工智能，我们去年推出了新加坡全国人工智能核心(AISG)。除了帮助企业，我们也必须培训我们的员工和国人，让他们有相关的技能，在工作和日常生活中有信心采用人工智能科技。
19. 追根究底，我们追求科技发展的 一大理念是科技应该以人为本。科技发展是为了让日常生活更便利，为各种挑战寻找解决方案，最终建设更美好的生活。在积极开发新兴科技的同时，我们要牢记科技以人为本的初衷。