政务部长Jasmin Lau在AWS新加坡公共部门日上的开幕致辞

Mr Jeff Johnson, Managing Director, ASEAN, AWS;

Ms Elsie Tan, Country Manager, Singapore Public Sector, AWS;

Distinguished guests, and fellow colleagues.

Good morning. I am delighted to join you at this year’s Public Sector Day, organised by Amazon Web Services and GovInsider.

Before we dive in, let me do a quick survey – how many of you have attended at least five AI conferences this year? (If you are seated next to someone who put up his hand, well, that’s the expert you should be listening to, not me!) But I'm glad most of you are still relatively fresh to this conference circuit, and therefore where you are today should give you a good flavour of what is to come and the possibilities out there that all of us can take on.

Think of today’s gathering as less of a conference or event, but more as a bridge between two ideals – our shared vision to reimagine public service, and our conviction that technology must serve every Singaporean's daily life. The real challenge is building the bridge to get us there, and no single organisation can build that bridge alone.

This is why gatherings like these matter -- when our public and private sectors come together, we don't just talk about transformation. We make it happen, one citizen at a time, one service at a time.

Let me share some stories about our efforts so far, and I hope that they will inspire you, as we continue on this important mission.

The first bridge we must build, is the most personal one. Transformation must begin from within – within ourselves, and our immediate teams. Until we ourselves actually pick up the tools and use them in our daily work, they remain just promises on a PowerPoint slide, or a Canva slide.

Everybody is talking about AI. Sometimes I do want to ask, so do you use AI, or do you just talk about it? Because once you start to use the tools, you start to realise that it's not at this point as great as you think it could be. You have to start using ChatGPT, or Pair, before you can realise the full potential and the downsides of it. Only when we get our hands dirty, as leaders, do we discover what is truly possible, and what is not.

I’ll share with you another example. When I took part in the general elections, we were all told that we’d better up our social media game. I didn’t know what kind of tools social media managers used to make these fancy reels. I realised that being a candidate for the election meant that there is no team to help you. You have to go figure it out yourself, including going to learn CapCut. I found out that no matter how good the tool is, if the subject is not funny or engaging, no amount of video editing is going to get you anywhere.

So it's really about getting our hands dirty, learning how to use these tools, learning what it cannot do, and I think that's the way we can bring out the best of these tools.

That is why, as DPM Gan announced recently, every public officer will take a mandatory AI literacy course. This builds on our efforts over the past year, where our senior Government leaders rolled up their sleeves to learn everything, from how the internet works, to how to build a digital product or a web application. Because we cannot ask our people to embrace what we do not understand ourselves.

Learning to us is not a luxury. Learning is how we stay relevant to the people we serve. It’s a responsibility that we have to take seriously.

I was particularly inspired, hearing about the LAUNCH! AI Incubator programme, where public officers, many without any technical background, built AI prototypes to solve real work challenges. With AWS support, teams from ITE and Workforce Singapore created tools for automated marking and job matching. These were not tech experts – they were everyday public servants who saw problems and decided to build solutions. As leaders, we must create space for our people to tinker and learn. That's how transformation grows from the ground up, not from the top down.

As we experiment with AI, we realise that technology is not a magic wand – it is a key that unlocks only the doors that it was designed for. Understanding the problem – really understanding it – is what is going to let us use technology responsibly and effectively.

Now, many of us in our agencies have technology teams and engineering departments. If any of the officers there do not understand the problem they are trying to solve, or have to ask their boss on a daily basis, “what do you want me to do?”, we are in trouble. We need to help a lot more of our public officers own the problems, fully understand it, know the groups of citizens that we are designing the solutions for, then there's hope that we get something good out of our systems.

As public officers, and it was not too long ago that I was a civil servant – we spend countless hours writing briefs, synthesising policies, and preparing presentations. I remember those late nights, with 40 tabs open on my laptop, drowning in the complex information, data points and policy papers, trying to distill complex information into something clear and actionable.

And because we knew this pain so intimately, we immediately recognised what language models could do for us, and this is why today we have government AI tools like Pair and AIBots, and these are now our daily companions. They don't replace our thinking, but free us to focus on the thinking that really matters. I don't know if we have young public officers here who today already benefit from AIBots or chatbots to answer public queries.

In the past, there were many of us whose jobs were to write the FAQs for websites so that when people Googled the FAQ, the correct answer will pop up. People used to spend days, if not weeks, just writing hundreds of FAQs, then it has to go through multiple rounds of vetting. Today, because of AI, a lot of that work is no longer necessary. Many more of us can now spend our time actually engaging citizens and supporting those whose questions may not fit into the general list that many people seek answers from.

You will hear the Pair team share their journey later, showing how they tailored AWS's generative AI services for government work. I also want to thank the Pair team personally, because I use Pair now at least twice a week.

This teaches us why deep domain knowledge and a good understanding of the user's pains is what separates responsible AI use from reckless experimentation. We need more of what I call “bilingual” practitioners, people who speak both the language of AI and the language of the field that they are in, be it education, healthcare or urban planning. We need more of these officers, more of these teams, to be able to develop AI effectively for the public service.

Partnerships can give us more of these bilingual teams. A great example is Health Kaki, which I believe is showcased today. MOH, HPB, Synapxe, AWS, and Temus collaborated to create a digital health companion. It now translates health goals into personalised daily nudges – healthier recipes, exercise tips – using AI to put behavioural medicine into practice. You can check out the Health Kaki booth at the exhibition outside later!

But let’s be honest about the limits of AI.

The moment we start using these tools ourselves, we see them clearly – the hallucinations, the outputs that miss the mark, the chatbots that cannot truly empathise or understand nuances. I’m sure there are lots of you who enjoy limit-testing and jail-breaking the AI to find out what kind of funny things it will do.

This is why we must never forget: we have the power to shape technology, ensuring it serves and strengthens us, not the other way around.

I spoke in my maiden speech in Parliament last week about my own fears that over time, technology and AI could become our digital opium. I don't know if some of you caught it, but it is something I feel very strongly about. At the start, it's fun, at the start it's helpful, and after a while, we become lazy. We find out that what used to take two hours to do, now I can take 10 minutes, and my boss doesn’t catch it. My boss doesn't seem to be reading it either. And then that's how they get addicted.

That's how we lose the humanity. That's how we stop putting in wisdom into what we do. We're very early in the game, not there yet, but I personally am very conscious that we could get there. Therefore I encourage all of you here to pay a little bit more attention to that. Yes, our lives are harsh. Yes, AI will help us. But remember not to use humanity when we try and use a little bit more technology.

That's how we lose the humanity. That's how we stop putting in wisdom into what we do. We're very early in the game, not there yet, but I personally am very conscious that we could get there. Therefore, I encourage all of you here to pay a little bit more attention to that. Yes, our lives are harsh. Yes, AI will help us. But remember not to lose humanity when we try and use a little bit more technology.

So this means holding on firmly to the values that make us human, and designing every tool, every workflow, every system with intentional space for human insight where it matters most.

We see this quite often in service delivery in the public service. The Government is exploring using AI agents to guide clients through simple administrative queries, so frontline workers can focus on counselling and advice.

Similarly, CPF’s call centres use AI to identify callers who struggle with self-help options, redirecting the less tech-savvy citizens to actual human officers who can provide human care, assistance and support.

Ultimately, machines will and do process information faster than all of us, but they cannot counsel a family in distress, affirm a struggling student, or advise a business owner through a tough decision.

Finally, our efforts reach far beyond transforming public services. They will inspire companies and citizens to join this journey. The more, we, in government, understand technology, the more we understand what support, enablers, and mindsets our businesses and workers need to thrive in the digital economy.

This focus on enablers also guides my work on the Economic Strategy Review Committee. I chair the Committee on Technology and Innovation. It’s been a good few months of understanding how businesses in different domains use AI differently. When I speak with companies to understand their challenges, one anxiety often surfaces – technology is racing ahead, while our skills and mindsets struggle to keep pace. So this is a gap that we will have to figure out as a country: how do we ensure that our workforce and our workers have the right skills and the right mindsets take on these challenges ahead?

That’s why I’m pleased to announce updates to AWS’s AI Spring initiative, with two new MOUs signed between AWS and Institutes of Higher Learning.

Firstly, through the new AI Nexus Lab, Nanyang Polytechnic students will have the chance to help prototype solutions for SMEs, and work with AWS platforms and partners to scale these solutions.

Firstly, through the new AI Nexus Lab, Nanyang Polytechnic students will have the chance to help prototype solutions for SMEs, and work with AWS platforms and partners to scale these solutions.

Secondly, Republic Polytechnic students will learn to use AWS's agentic coding platform, Kiro, as part of their curriculum – honing practical skills for the future of work.

Such efforts bring us closer to our vision of an empowered workforce – one that can apply AI thoughtfully and critically in their respective fields.

Let’s continue to work together to empower every company and every employee, so that the benefits of AI reach and uplift all Singaporeans. The AI wave is here, and it’s already reshaping the way we live, work, and serve. But with every wave comes a choice – to be swept along, or to rise and ride it purposefully.

As a Government, we must help ourselves and our people embrace this technology with clarity and confidence. This journey can only succeed if we work together – to understand AI and emerging technologies, apply them with purpose and principle, and empower every Singaporean to do the same. That is how we bridge the gap between what technology can do, and what our people need it to do.

Thank you, and here's to learning and building together! I wish you all a good day ahead!