Everything you need to know about Singapore’s AI Strategy
Our Smart Nation: The beginning of the next 50 years
During the 2015 National Day Parade, we looked back on the 50 years of achievements that transformed Singapore into the bustling metropolis it is today. With that narrative in mind, many Singaporeans began to wonder what the next 50 years would look like and whether this little red dot could continue to remain exceptional — a country capable of punching above its own weight.
The Smart Nation initiative was released shortly after to address these questions. It became clear that we needed to harness the power of technology to remain competitive as a global city and to improve the lives of our people. Some of us thought that this vision was uninspiring and vague at best, as the first-tier cities in China seemed more advanced than us despite not calling themselves a ‘Smart Nation’.
Perhaps we were used to consuming well-crafted responses and expected the government to have all the answers as they steered us through these unchartered territories. It was only recently when the Smart Nation Digital Government Office (SNDGO) decided to provide more information about its plans did everything begin to fall in place. The government has decided to launch several projects including building a National Digital Identity (NDI) among other initiatives in a bid to build a government, an economy and most importantly a society that is digitally enabled.
Entering the next phase of our Smart Nation narrative
Fast forward to today, with Artificial Intelligence (AI) all the rage, it is almost fashionable nowadays to have an ‘AI strategy’ to share with the world. However, Singapore’s strategy is deeper and broader than just a token white-paper. At the 2019 SFF X SWITCH conference, Singapore released her National AI Strategy (NAIS) that provides concrete action plans and timelines that detail how AI is the core of bringing us into the next phase of our Smart Nation journey.
The strategy also outlines how Singapore intends to engage with other countries, both private and public organisations as well as individuals in writing the next chapter. In order to co-ordinate the efforts across the various agencies to prevent each fraction from working in Silos, the new National AI Office (NAIO) was created to steer these efforts in the same direction.
Despite aspiring to become a leader in developing and deploying scalable and impactful AI solutions by 2030, it is clear that Singapore places a strong emphasis on being able to deploy rather than develop AI solutions. Being a small nation with limited resources and capabilities, we have chosen to leverage on what differentiates us. Prime Minister Lee Hsien Loong mentioned at the recent World Economic Forum (WEF) that despite the growing anxiety and mutual suspicions between the 2 great superpowers, it would be foolish for us to strategize expecting to grow our own food and produce our own computers because we won’t survive under such conditions.
As such, the nation has decided to leverage on our superb infrastructure for technology deployment, availability and access to a rich and diverse corpus of data, a large pool of highly educated and trained professionals coupled with a politically neutral climate that has a strong rule of law to attract the scarce talents and investments in this competitive AI race.
Building an ecosystem to enable deployment
To facilitate the testing and scaling of such solutions, the strategy outlines five ‘Ecosystem Enablers’ to drive AI innovation and adoption across the economy.
- Triple Helix Partnership
Objective:
Enable rapid commercialisation of fundamental AI research and deployment of AI solutions.
Key thrusts:
1.1. Deepen investments in AI-related R&D across the ecosystem
Following in the footsteps of how Canada became the world’s leading AI research hub, Singapore is looking to invest more in AI-related research after having committed S$500m under the RIE2020 plan. Our supercomputing facilities will also be receiving an additional S$200m worth of upgrades to bring its performance up from 1 petaflop to 15–20 petaflops to support the development and training of new AI algorithms and models. Moreover, it seems that we might have a slight head start as Singapore was recently ranked first in the world based on the impact of our AI scientific publications according to Elsevier SciVal in 2019, with four of the ten most promising young AI researchers in 2018 currently based in Singapore.
1.2. Drive partnerships between the research community and industry
Currently, there are more than 15 AI public-private partnerships and joint labs in Singapore to support the commercialisation of AI technologies. Apart from enabling the research community to gain access to relevant data and sharpen their focus on market needs, partnerships like the Singtel-NTU AI Lab hopes to benefit industry with these research innovations and having access to the pipeline of research talent.
1.3. Accelerate AI adoption in companies
In order to include the majority of Small and Medium Enterprises (SMEs) in this digital transformation, there are numerous initiatives to support them on their AI journey such as SMEs Go digital programme, AI Singapore’s 100 Experiments programme, AI Makerspace, Accreditation@SGD and SG:D Spark programmes, Tech Depot on SME Portal as well as IMDA’s Digital Services Lab.
1.4. Establish AI innovation testbeds
By 2023, the Punggol Digital District (PDD) would be a living testbed that will integrate various technology verticals including facilities management systems, district cooling systems, autonomous goods delivery systems, and security systems.
The Singapore Institute of Technology (SIT) and JTC business park would also be co-located in the PDD to facilitate close collaboration between the digital industry and academia in hopes of transforming what it means to work, learn, live and play in a smart district.
2. AI talent and education
Objective:
Address the shortfall in the quantity and quality of talent across the entire range of AI-related job roles.
Key thrusts:
2.1. Train more Singaporeans for higher value-added AI jobs
It is crucial that a country establishes a local talent pipeline to ensure the sustainability of her national initiatives. Therefore, Singapore intends to train more AI PHDs with industry experience through postgraduate scholarships with companies such as Alibaba, Nvidia, SenseTime and Grab.
In addition, other programmes including the AI Apprenticeship Programme (AIAP) and the TechSkills Accelerator (TeSA) offer Singaporeans opportunities to acquire new skills and competencies in AI by working on real-world industry projects.
2.2. Teach basic computing skills and computational thinking to all
Developing basic AI competencies and literacy in students by sparking their interests at a young age will help the nation in developing ‘bilingual individuals’ who can apply computer science and AI concepts in their respective domains.
Tertiary institutions are currently introducing domain-specific AI courses to ensure undergraduates understand how to leverage on AI tools in their work. Moreover, AI literacy courses for adult Singaporeans and school-going children are also being conducted to bring them along this journey.
2.3. Attract top-tier global AI talent
Apart from building local talent pools, growing our global talent networks and remaining open to top-tier AI talent will increase the chances of Singapore being exposed to the latest cutting-edge technology.
To facilitate this, the Economic Development Board (EDB) and Enterprise Singapore (ESG) will pilot the Tech@SG Programme that fast tracks employment pass applications for core team members in fast-growing technology companies who are equipped with deep technical skills.
3. Data architecture
Objective:
Enable quick and secure access to high-quality, cross-sectoral datasets.
Key thrusts:
3.1. Establish frameworks for public-private data collaboration
Data is one of the key ingredients for training and scaling many AI solutions, and Singapore has the unique ability to provide timely (and secure) access to rich, high-quality datasets. Given our heterogeneous and multi-ethnic population, companies can leverage on our diversity to develop exportable AI solutions in sectors such as healthcare that could be scalable across Asia.
In order to manage our data as strategic national assets and establish an ecosystem to acquire, clean, fuse and distribute data, the Infocomm Media Development Authority (IMDA) has recently published the Trusted Data Sharing Framework to guide companies in establishing data-sharing partnerships with each other, pivoting away from the Personal Data Protection Commission’s (PDPC) previous position of discouraging entities from collecting and sharing data. The framework includes the key legal, regulatory, and technical considerations and safeguards to provide greater clarity on the scope and governance for sharing data across public and private sectors.
3.2. Establish trusted data intermediaries for public-private data exchange
As the nation’s custodian of personal and administrative data, the government intends to work with companies in key sectors to define and communicate a set of common data standards for the sector to ensure data interoperability. This will help facilitate data discovery and AI collaboration between Government agencies, industry, and researchers.
4. Progressive and trusted environment
Objective:
Strengthen trust in AI technologies to enable an environment for test-bedding, developing, and deploying AI solutions.
Key thrusts:
4.1. Establish citizens’ trust on the responsible use of AI
As AI becomes more pervasive, we need to shape how it interacts with citizens, institutions and society. The plan outlines efforts to curate AI solutions that are explainable, develop codes of conduct and training in AI ethics and governance for professionals managing AI solutions and implementing AI projects as well as to work with institutions such as the Singapore Management University’s Centre for AI and Data Governance to explore the potential options of addressing longer-term issues.
4.2. Provide a top-class IP regime and accelerated patent initiatives
Further solidifying our position as a safe haven that enforces strong intellectual property rights to encourage AI innovators to commercialise new products and services, the Intellectual Property Office of Singapore (IPOS) has launched the Accelerated Initiative for Artificial Intelligence (AI2) which is said to be the world’s fastest patent-acceleration programme that grants applicants an AI patent within just 6 months.
5. International collaboration
Objective:
Work with international partners to shape the international AI discourse and develop the other horizontal enablers.
Key thrusts:
5.1. Contribute to global standards for AI-related policies and guidelines
Singapore will continue to work with key international organisations such as the World Economic Forum (WEF) and the Organisation for Economic Co-operation and Development (OECD) in a bid to build thought leadership in setting AI-related policies and standards so as to remain relevant in this rapidly evolving field.
5.2. Collaborate on multi-national AI projects
AI Singapore has established ties with serval countries to catalyse joint AI research pursuits in hopes of bringing together an international network of multi-site testbeds for AI solutions. One such collaboration was formed in July 2008 between the French Institute for Research in Computer Science and Automation (INRIA), AI Singapore, the French National Centre for Scientific Research (CNRS), and the French Institute of Health and Medical Research (INSERM) to focus research on AI in healthcare, explainable AI, federated and distributed learning, natural language processing and privacy, trust and accountability in AI and data sharing.
Anchoring talent and investment through the 5 National AI Projects
Internally, the strategy is set to tackle 5 key national challenges for Singapore in transport and logistics, smart cities and estates, healthcare, education, and safety and security. These projects were identified based on their high social and/or economic impact and it serves as a guide for how Singapore intends to invest in its digital infrastructure moving forward.
At our recent panel discussion with the senior assistant director, Mr Huang JueJing, of the new National Artificial Intelligence Office, he said that in order to attract the best talents and companies to Singapore, there must be opportunities for them to work on challenging and impactful problems. Singapore needs to bring together rich datasets for research and create an environment where the best minds can push the boundaries of science in a safe and ethical manner.
Remaining relevant in an increasingly uncertain future
The race for AI supremacy is on and the release of this roadmap could not have been timelier. With the rise of geopolitical tensions in the region, it might be prudent to explore the possibility of reducing our vulnerable dependence on imports of international technology. As we become increasingly dependent on technology to maintain our competitive advantage, we should take measures to ensure a reasonable level of self-sufficiency in producing certain crucial hardware (i.e. AI chips). Singapore needs to continue developing competencies to remain relevant so as to secure a seat at the table or we might very well be on the menu, having our fates decided in our absence.
Written by: Sheldon Chin (28/1/20)
Edited by: Jeremy Ang (28/1/20)
