Summarized — National AI Strategy (NAIS) 2.0, and the Actionables
The National AI Strategy 2.0 has been published!
Or to be more precise, it has been released for a few months now 😅. Named NAIS 2.0, the paper with the same name depicts the ambitions of Singapore in the field of AI in the next few years. Why NAIS 2.0? Because NAIS 1.0 was released in 2019.
Unfortunately, NAIS 2.0 published paper has a breathtaking amount of content, being 68 pages long!
Lucky for you 😬, I’m here to grind my teeth and read this entire paper, summarizing and adding some of my interpretations of this paper 😏. Do take note of the “15 actions”, something that should concern those who are interested in, or at least just observing the situation in Singapore.
Let's start to explore this paper — NAIS 2.0, for Singapore!
Disclaimer: This is entirely read by me and regurgitated mostly in my own words. I didn’t cheat and use Gen AI tools like Bard or ChatGPT, so no meta-joke from me on AI summarizing AI strategy here 😬.
Also, as a Singaporean, I know we love our acronyms and abbreviations, so be prepared to see plenty of those 😁.
Finally, there’s a summary of this summary at the end of this article, so if you want to just get to the point, simply scroll all the way down!
Paper — NAIS 2.0
Note that in the following sections, I’ll do my best to summarize the content into bullet points, while adding (with indication) my interpretation of what some of the abstracted content could mean.
Section 1 — Foreword
- Mostly talks about Singapore’s recognition of the rising trends in the prominence of AI in 2023, and our Government’s aspirations to develop in this area.
- Using a quote from the report here that sums up the entire paper quite succinctly —
“… plans to build a thriving AI ecosystem; develop our workforce to take on new opportunities; provide enough infrastructural capacity to achieve our ambitions; and foster a trusted environment that protects users and facilitates innovation.”
Section 2 — Introduction
- Refers to AI as strategic importance, for the “next frontier of economic growth” for Singapore.
- Another interesting note in the introduction — “Address risks from the potential abuse and mismanagement of AI”.
- (Interpretation) This seems to reflect the current landscape of Singapore afflicted by scams and focuses more on computer ethics (human use of computers), rather than machine ethics (morality of AI).
Within the introduction, it also touched on some interesting statistics:
- (On funding) “We committed more than S$500 million through AI Singapore (AISG) under the Research, Innovation and Enterprise (RIE) 2020 and 2025 plans”
- (On ranking between nations for AI support) “Singapore is … ranking among the top 10 in the world. Over 80 active AI research faculty, 150 AI R&D and product teams, and 1,100 AI start-ups call Singapore home”
Section 3 — National AI Strategy 2.0
This section briefly touches on “Vision and Goals”, as well as “Plans”, which is the main meat of the entire report.
On “Vision and Goals”:
- Twin goals of achieving excellence and empowerment surrounding AI, focusing on manpower, businesses, and communities.
On “Plans”, the following image is used, which is delved into deeper in the later sections (don’t over-index on it as it will be explained later on):
The key point here is “3 Systems” to power Singapore’s AI aspirations:
- Activity Drivers
- People & Communities
- Infrastructure and Environment
Under each of the systems, “10 enablers” are listed which is further distilled down to “15 actions”.
These “15 actions” are what I considered to be the most interesting for this paper, which are the “actionables” committed to over the next 3–5 years.
Section 4 — “System 1, Activity Drivers”
As the name suggests, the first of the 3 systems, “Activity Drivers”, aims to power the growth of AI, through the push of increased activities, with enablers “Research”, “Government”, and “Industry”.
The actionables are best summed up in the first 4 of the “15 actions”:
Action 1 — Anchor new AI Centres of Excellence (CoEs) across companies, and explore establishing Sectoral AI CoEs to drive sophisticated AI value creation and usage in key sectors
- Anchor and attract AI producers and end users (startups and tech companies).
Action 2 — Strengthen our AI start-up ecosystem, including attracting AI-focused accelerator programmes to spur rapid AI experimentation
- Attract VCs and nurture accelerator programs to encourage the growth of AI innovators and related intellectual property (Note — no specific numbers offered here in this paper).
Action 3 — Improve Public Service productivity, with new value propositions for our citizens
- Priorities of “Smart Nation”— Implement AI strategies to guide and resolve challenges faced within the public domains (e.g. Healthcare, Education).
- For government agencies that lead functional domains (e.g. Finance), to work with counterparts in the private sector that have the potential to use AI and move towards streamlining their business use cases.
- (Sidenote) An excerpt here does mention that one focus will be on uplifting public officers’ capabilities in understanding and working with AI. This is reassuring to ensure that policymakers are committed to understanding what they are working with.
Action 4— Update national AI R&D plans to sustain leadership in select research areas
- 5 focus R&D areas — Research Priorities, Industry-Academia Nexus, Talent, Compute, and International Collaborations.
- Notable point 1 — Government to be a present facilitator of more R&D collaboration, between academia and the private sector.
- Notable point 2 — Strengthen efforts to recruit top AI researchers within and outside of Singapore (Action 5 will touch on this more).
- Notable point 3 — Secure key resources, like GPUs, for the research community.
- (Interpretation) Possibly related news: CNA: Nvidia could invest in ‘iconic’ AI site in Singapore: CEO Jensen Huang)
Section 5 — “System 2, People & Communities”
Following on, we have the next section, which touches on System 2 — “People and Communities”. This portion is interesting — focusing on increasing the AI talent pool, as well as the possibility of a centralized physical hub for AI in Singapore.
The enablers here are “Talent”, “Capabilities”, and “Placemaking”, driving the next 5 of the 15 actions to be taken:
Action 5 — Attract world’s top AI Creators to work from and with Singapore
- A new team in Singapore (assumed Government agency but not specified in this paper) is to be set up, for engaging and attracting global AI talents. This team will also serve as a platform for talents and companies to integrate into Singapore.
- Parallel efforts will also be focused on inducting AI experts and institutes into our ecosystem without physical presence — such as through hybrid working arrangements and partnerships with global institutes.
Action 6 — Boost AI Practitioner pool to 15,000
Before we jump into the bullet points, this is perhaps the “magnum opus” of the paper, quoting an actual number of 15,000 AI practitioners to be achieved in the next 3–5 years, and publicized by major media outlets.
Ok, so here’s the summarized view of this action:
- Scaling and refining the AI Apprenticeship Programme (AIAP), a 9-month, full-time apprenticeship program that includes company attachments to bolster the participant’s experiences.
- Scaling up technology and AI talent pipelines, focusing on pre-employment training.
- (Interpretation) The term “pre-employment” is unclear, but my best guess is that it refers to Universities, Polytechnics, and technical institutes. - Remain open to the global talent pool (as covered in Action 5 too).
Action 7 — Intensify enterprise AI adoption for industry transformation
- Prepare tools for companies to use, for evaluating readiness to adopt AI — such as the AI Readiness Index (AIRI).
- The government will also utilize programmes like CTO-as-a-Service (CTOaaS) and Digital Leaders Programme (DLP), to guide companies in properly transitioning to using AI, and also to build internal capabilities for AI.
Action 8 — Upskill workforce through sector-specific AI training programmes
- This action is likely an overlap with “Action 6” above, but it also focuses on a appointed “Sector lead” for different sectors (e.g. Financial Services sector) to drive sector-specific AI upskilling programs
- (Interpretation) “Sector lead” is an ambiguous term that doesn’t touch on the origin of the personnel — whether public or private. My best guess here is that it refers to some government agent.
Action 9 — Establish an iconic AI site to co-locate AI creators and practitioners, and nurture the AI community in Singapore
- Setting up a dedicated physical place for AI in Singapore. While the actual location is not specified, the paper compares it against SF as another AI Hub.
Section 6— “System 3, Infrastructure and Environment”
I hope you are still alive 😆. Stay with me here, we’re on the final section covering the 3 systems — Infrastructure and Environment!
This system revolves around base capabilities brought about by needed infrastructure to drive AI innovation, as well as the environment for empowering AI development.
It is served primarily by 4 enablers — “Compute”, “Data”, Trusted Environment”, and “Leader in Thought and Action”. The actions accompanying the system and its enablers comprise the final 6 of the 15 actions:
Action 10 — Significantly increase high-performance compute available in Singapore
- Support data centers with GPU capabilities, ensuring a sufficient carbon budget and power allocation on a regulatory level. In the long term, carbon-neutral, green data centers will be essential.
- (Interpretation) Early on, possibly concessions for national policies like carbon tax and energy allocation, while additional support to transit to renewable energy sources in the longer term. - Deepen partnerships with chipmakers and cloud providers, ensuring local access alongside these partnerships.
- Interestingly, the paper also touches on the Government managing its own GPU clusters for “meritorious use cases”.
- (Interpretation) Possibly hinting at a government-run entity for AI cloud computing, for the sole purpose of serving the public good?
Action 11 — Build capabilities in data services and Privacy-Enhancing Technologies (PETs)
- Develop deeper capabilities in Singapore for PETs, through expanding measures for sandboxes and managing guidelines on experimenting with PETs, thus expanding data that can be used for insights (unclear on the source of data, but perhaps referring to Government collected data that will be properly treated before sharing)
- An example is the PET sandbox, released by IMDA in 2022, for providers to trial and build use cases with PETs
Action 12 — Unlock Government data for use cases that serve the Public Good
- Split into 2 areas, “Public” and “Private”
- For “Public”, the government is to selectively share closed datasets, through agreements restricted to the core theme of “Public Good” (not referring to the noun for goods, but referring to the benefit of the public)
- For “Private”, the government would selectively act as an intermediary role between different private entities in sharing and aggregating data.
Action 13 — Ensure fit-for-purpose regulatory environment for AI
- Overall, this action outlines the Singapore Government’s outlook on being involved as partners within the AI ecosystem, and also within the international AI communities, to create reasonable safeguards and broader legislations for varying use cases.
Action 14 — Raise security and resilience baseline for AI
- This action is split into 2 areas — Short term and Long term
- Short term — Work with partners to drive secure AI adoption
- Long term — Establish and rollout best practices and guidelines to improve AI security
Action 15 — Establish Singapore as an ambitious and pragmatic international partner on AI innovation and governance
- In this action, the paper reinforces the Singapore Government’s commitment to being a “serious and reliable international partner on AI”.
- Firstly, local engagements with global entities on topics surrounding AI by the Government, such as governance frameworks like AI Verify (which is wholly owned by IMDA):
- Next, international engagements are also a priority. One notable mention is that Singapore is the convenor of the Forum of Small States (FOSS), which puts her in a position to collaborate and lead AI development and initiatives with its 108 members.
Section 7 — Final section — A Whole-of-Nation Movement
We’re finally at the end! The final section here is only a page long — and covered in its entirety within the above.
Similar to themes in the foreword and introduction, it touches on the importance of AI on economic growth and competitiveness, as well as acknowledging and inviting local and global entities who resonate with the intentions of the paper to form partnerships with and within Singapore.
Summary of the summary 😆
After all the above, here’s a summary of this summary post by WeiyuanGPT (yours truly).
- Action 6, as I mentioned — is likely the “magnum opus” for both individuals intending to venture into this sector, and AI entities planning to set up shop in Singapore to tap into her AI talent pool.
- For working adults who are interested in reskilling into this sector, AIAP is going to be something you want to look out for.
- Note that recent news articles characterize this policy as “tripling the workforce”, which if interpreted along those lines, means that a pool of 10,000 experts and jobs will be created in the targetted in next 3–5 years, in addition to a pool of 5,000 (interpreted) that we already have today.
- However, there seem to be some inconsistencies in recent news articles’ mention of “tripling the workforce”. The NAIS 2.0 paper mentions only the number of institutes and teams, with no indication of the local talent count. As such, we might want to observe further and take the statement of “tripling” with caution for now:
- For entrepreneurs and students, there would be more government and private support towards locally spun-up AI initiatives, perhaps reminiscent of the startup wave in the early to mid-2010s. More observation is needed here until plans are unveiled on government support along with encouraging VCs in this domain.
- For international and local researchers, and also research institutes — the proposed team to serve as the platform for engagement in AI research is one to watch, although no timeline has been given yet. Reservation of computing capabilities and promised stronger ties with providers and data centers, are reassuring as a commitment for Singapore to be a center of AI excellence and empowerment.
- For grassroots AI interest groups and organizers, there is some promise in the form of the physical infrastructure (i.e. a centralized district for AI entities) for driving interest and forums in future events. As a local grassroots leader in similar interest groups, it would have been great for more written plans to support these groups beyond infrastructure, but worth observing how this develops into the future.
- For all citizens, there would be stronger dependence on AI for day-to-day needs — this might be something to observe in our everyday lives (such as in domains like healthcare and education), interesting to see how this might play out.
- An overall read of all of the indicated 15 actions does seem to display commitment and understanding by the Government to drive at all levels of AI innovation — from manpower, infrastructure, legislation, to even methodology focus (e.g. PET, and also governance frameworks). Recognition of both national and international efforts from the start was also a good callout to work on.
- More numbers would have been great, but this is also a good start for Singapore, in her dedication to being a world leader in AI and creating confidence to drive towards increasing the talent pool while matching it with the equivalent job openings.
Yup that’s the end of this article!
.
.
.
.
.
.
Wait, you want more?
Well, here’s something more — a list of all those acronyms and abbreviations 😬
- NAIS — National AI Strategy
- AISG — AI Singapore
- RIE — Research, Innovation and Enterprise
- CoEs — Centres of Excellence
- AIAP — AI Apprenticeship Programme
- AIRI — AI Readiness Index
- PET — Privacy Enhancing Technology
- CTOaaS — CTO-as-a-Service
- DLP — Digital Leaders Programme
- IMDA — Infocomm Media Development Authority
- FOSS — Forum of Small States
