Speech by Deputy Prime Minister Lawrence Wong at SCAI

Speech by Mr Lawrence Wong, Deputy Prime Minister and Minister for Finance, at the Singapore Conference on AI for the Global Good to Launch Singapore National AI Strategy 2.0

4 December 2023

My Cabinet colleague,
Minister Josephine Teo,
Excellences and distinguished guests,
Ladies and gentlemen,

I am happy to join you this evening for our inaugural Singapore Conference on AI for the Global Good. Let me also extend a warm welcome to all our overseas guests at this beautiful venue, the National Gallery of Singapore. It is quite interesting we are discussing cutting edge science here in the gallery of the arts. But it is a wonderful synthesis of both science and arts and bringing every one of you together.

This past year has been transformative for AI, as I am sure all of you know.

• AI technologies have been around for decades. Veterans in the field, including many of you in this room, would have experienced several booms and winters; multiple phases of advancements and stagnations.

• Yet recent breakthroughs in AI have sparked renewed interest – about the potential of AI, its risks, and its implications for humanity.

• Regardless of our individual perspectives on these issues, one thing is clear – AI now stands as a pivotal game-changing technology that impacts us all.

Discussions about AI’s potential benefits and threats are also not new. But we are venturing into unchartered territory.

• Up to now, AI has been mainly about pattern recognition.
• And we are getting better and better prediction machines.
• But in time to come, we will have AI systems with agency and with transactional abilities, we will have machines with human-level cognitive abilities and the capacity for self-awareness and independent decision making.
• And all this will fundamentally reshape our way of life, with profound implications for societies everywhere.

In Greek mythology, Prometheus bestowed humanity the gift of fire. It was a tool that humanity successfully mastered to survive and thrive in a hostile world. Likewise, we must master AI to wield its potential for our collective benefit.

Singapore has been taking steps to master and make good use of AI.

• In healthcare, we are using it for disease detection and diagnosis.

• In finance, we are using it for fraud detection and risk management amongst other things.

• In the public service of Singapore, we have embraced AI.

  i. We use large learning models to help our officers reply to public queries. It has been very helpful for them, because with a click of a button you get a draft prepared for you, and most of the work is already done.

  ii. We are using our own secure version of ChatGPT to improve productivity. For example, for coding and writing assignments. AI has not helped me to write my speech yet, I still have to do it on my own. But I am sure if I put this through ChatGPT it will give it an excellent summary, and in time to come AI will be able to draft a fairly decent speech.

It is not by chance that we can do these in Singapore today.

• In 2019, we were amongst the first countries in the world to publish a National AI Strategy.
• We were also the first country in Asia to establish a Model AI Governance Framework that articulates guiding principles to develop AI that is explainable, fair, and human-centric. We have a regulatory environment that supports AI experimentation and innovation.
• And our researchers in our universities are well-regarded internationally, with expertise in natural language processing and computer vision.
• All these have all contributed to our ability to harness AI for economic gains and social benefit.

With recent advances, especially in Generative AI, we have seen the spectrum of possibilities expanding tremendously.

• Our vision is for Singapore to be a place where AI is used to uplift and empower our people and our businesses.
• We want to harness AI for the public good, for Singapore and for the World.

This revised strategy is the result of extensive consultations with over 300 experts and organisations, both domestic and international. Many of you in this room contributed to our refreshed strategy, so we want to extend our heartfelt appreciation and gratitude to all of you. Thank you very much.

In our refreshed strategy, we will aim high, we will dream big, and we will set ambitious goals. We will foster a spirit of boldness, experimentation and innovation in our next phase.

At the same time, our ambitions are rooted in reality. We are acutely aware that every leading city in the world today wants to be an AI hub. The competition is fierce and will only get tougher.

Singapore is a tiny little island, with no natural resources. We cannot compete head-on in terms of size or fiscal resources. But we do have several factors in our favour: a highly-skilled workforce, a highly wired-up society; and, importantly, a trusted eco-system where things work, and where we can make things happen.

So our strategy involves active participation in AI research and deployment. We aim to scale up AI for compelling use cases in areas like advanced manufacturing, financial services, healthcare, education and public services, and bring benefits to others and to ourselves in Singapore.

In undertaking these national plans, we will work in close partnership with researchers, industry stakeholders and others. We want to foster closer collaboration with all of you. And let me just briefly highlight several areas that we are focusing on in Singapore.

One, we aim to build skilled AI talent in Singapore.

• We think of talent in three categories.
• The first is Creators. Top-tier talent engaged in novel and cutting-edge AI activities. We would like to recruit more of such top Creators, to work with Singapore and to work in Singapore. Some of you may be interested, if so, we would like to continue the conversations with you. The numbers involved will not be large. But we believe that these top Creators can have an outsized impact in fostering a culture of experimentation and excellence in AI.
• The second category is Practitioners. These are our data, machine learning scientists and engineers. They are the backbone of AI, and we aim to more than triple the number of AI Practitioners in Singapore to 15,000. We will train more locals who can excel in this space. We will also welcome foreign professionals with the right skillsetsto complement our local workforce.
• The third category are Users of AI-powered products and services. We aim to build up a base of confident AI users – both amongst enterprises and workers – who are equipped to leverage AI for enhanced productivity and for more impactful work.

We also recognise the concerns about the impact of AI on jobs and livelihoods.

• Knowledge-based work like research, coding and writing were considered safe from disruption in the past. But with AI, that is no longer the case.
• We do not think this will mean a jobless future. But it does mean significant changes in job roles, and more training for humans to harness AI effectively.
• That is why the Government plans to invest significantly in adult education and training, to reskill and upskill our workers.
• In short, we aim to nurture a critical mass of skilled AI talent; and at the same time, ensure that everyone in the workforce is AI-ready. So that is our talent strategy.

With people and with talent, we aim to build a thriving AI industry. And that is the second part of our strategy.

• We are already home to several AI start-ups. We have tech giants based here, we have MNCs with regional and global operations here as well. Many are leading-edge AI producers as well as sophisticated end-users.
• Today, many of them use AI on a project-basis, or for individual applications. This is a good start, but we want to do more in order to build a thriving AI industry in Singapore.
• In particular, we want more companies to establish their “AI Centres of Excellence” here in Singapore – undertaking value creation activities from model development to infrastructure building to application deployment.
• These centres can go beyond the needs of the specific company, and they can benefit the wider industry. There is potential for such industry-wide platforms, in key areas like advanced manufacturing, financial services, and biomedical sciences, where Singapore is already operating at the leading edge.

So we have talent, and industry. The third area is research.

• Indeed over the past five years, we have invested significantly in AI research.

• In the next bound, we will sharpen our focus to three research areas where we believe Singapore can develop a competitive advantage, namely in Responsible AI, Resource-Efficient AI, and Reasoning AI.

  i. Responsible AI because that will develop AI that is trustworthy and safe, reduce bias, increase reliability, and allow for auditability.

  ii. Resource-Efficient AI, because that will help us with the resourceintensiveness of AI on data, compute, energy, and carbon, all of which are pertinent to a small country like Singapore.

  iii. And Reasoning AI because that helps to improve the models’ understanding of logical and physical concepts, and explain their outputs, thereby expanding AI’s applicability

Of course, to achieve these goals, Singapore also needs reliable access to compute resources. We cannot compete with the major powers in assembling raw computing power. But we will do everything we can to ensure we have the computing power to meet our growing research and industry demands, and to fully back our strategic AI agenda.

Another area of action is to build a trusted environment.

• Even as we take steps to reap the benefits of AI, we are mindful of the risks and challenges.

• There are many ways that AI can be misused:

  i. A whole range of bad things that can happen, powered by AI: from deepfakes, to scams, to cyber-attacks and the spreading of misinformation.

  ii. There may even be problems with decision making: AI models may work well 99% of the time; but even that 1% can potentially wreak havoc and lead to disastrous consequences.

There are no easy answers or silver bullets to address this.

• If you just consider what happened with the internet wave.
• And we were just talking about this with a smaller group just now. Governments around the world largely adopted a laissez faire approach to social media. And now, everyone is paying the price.
• Countries are grappling with the consequences: cyber-bullying, the impact on mental health, especially amongst young people, or the spread of misinformation and fake news online.

With AI there will be even more profound issues to tackle.

• There will be moral and ethical issues.
• There will be wide-ranging implications for regulations and governance.

The temptation of course, is to swing to the other extreme and impose strict regulations now. But that is also not ideal.

• AI technology is very general, and its future uses are very hard to predict. At this stage, everyone is still figuring out what exactly are the best and most appropriate use cases for AI. It will not be possible for regulators to stay abreast of these changes, and figure out all the answers.
• And some overreach in regulations can also be counter-productive. For example, if we were to license AI systems on the condition that they prove they will cause no harm – I think that is very likely to stifle and to choke off all potential innovation.

So in the end, we will have to find a pragmatic balance. And that is precisely Singapore’s approach.

• We want to encourage experimentation and innovation.
• We will create regulatory sandboxes for such innovation to flourish here in Singapore and we will put in place the necessary guard rails so that innovation and experimentation does not cause systemic risks.
• As AI progresses and as the rate of scientific progress increases, we will continue to adapt and evolve our rules.
• The key in all this is to be agile and nimble, and to keep on updating our strategies and our governance frameworks as circumstances change. So that is our philosophy in Singapore.

Ideally, governance for AI ought to be global – because AI is decentralized and everywhere. You only need one model to fail, and it can have knock-on effects everywhere around the world.

That is why we also need to grow our international partnerships.

• It is not going to be easy to do so in today’s geopolitical environment.
• Where there is intensifying global competition in the AI arena. Countries increasingly see AI as a zero-sum game, which will give the winner a decisive strategic advantage.

We accept these realities.

• But we also need to ensure that competition does not distract companies away from the important questions on the direction of development of AI, and use of AI, that affects all of us.
• So the key is to bring together like-minded countries to collaborate and to shape the global governance architecture for responsible AI.
• Singapore will do our part in this shared endeavour.
• We will work with a wide range of international partners to set sensible “rules of the road” in AI, such as by enhancing interoperability between governance frameworks. We have done so with the US, and we look forward to similar engagements with other partners.
• We will develop research and technical collaborations to support innovation across borders, including through partnerships with the UK’s new AI Safety Institute.
• And we will continue to support multilateral, multi-stakeholder platforms. For example, at the WEF, the UN, or the Forum of Small States. Closer to home, we look forward to the launch of the ASEAN Guide on AI Governance and Ethics, during Singapore’s Chairmanship of the ASEAN Digital Ministers Meeting in February next year.
• Above all, we want Singapore to be a reliable and trusted partner on AI innovation and governance: where our efforts attract others to work with us, and earn us a place amongst the pace-setters in AI.

So, to conclude, I’ve shared several areas for action that reflect Singapore’s ambition in AI. Done well, we believe that AI can uplift our economic potential, enhance social impact, and in time, meet the needs and challenges of our time. And that is why we have brought together all of you in this conference, to identify the questions in AI that, if solved, can lead to the development and deployment of AI for the global good.

I hope that all of you will see this as a collective attempt to sow the seeds for the type of multi-stakeholder, multi-disciplinary, and multi-national partnership that a technology like AI demands of all of us.

So together, let us make a meaningful contribution to the advancement of AI for the global good. And on that note, I wish all of you a very fruitful and productive conference.

Thank you very much.