约瑟芬·谭部长在SGTech 2026慈善晚宴上的开幕致辞

SMS Tan Kiat How

SMS Zaqy Mohamad

Dr Yaacob Ibrahim

Dr Amy Khor

Our SGTech Patrons

Mr Nicholas Lee

SGTech Chairman

SGTech Board of Governors, members, colleagues and friends,

At the SCS Tech3 Forum last year, I proposed the idea of “AI bilinguals” – professionals who not only have strong domain expertise, but also a good command of AI.

Their fluency with two languages – their domain and AI – would make for a powerful combination.

Since then, this description has been borrowed and used many times over. Intuitively, Singaporeans understand what it means, because most of us are products of the bilingual education policy that has served Singapore so well.

However, as individuals and as parents, we know that learning a language is not easy.

For one, we are born with different gifts and talents; learning a language comes more easily to some than others.

Achieving fluency also takes effort. We will need a minimum vocabulary, which depends on the language we are learning. For example, to be fluent in English, we will need to know four to five thousand words to read and write. For Chinese, we will need about two-and-a-half thousand words.

Thereafter, we need to use the language regularly. Otherwise, we get rusty. The good news is that you never quite forget the language completely. With some help to practise, fluency in a language can be regained.

These features of language learning may well apply to AI. And so, our first hope is that everyone in the workforce gains basic literacy, and can achieve the equivalent of speaking a language sufficiently to communicate. Beyond that, we believe there are many who can become fluent in the use of AI, to do the equivalent of reading and writing to get more things done. Among them, we have a special interest in professionals.

Their fluency in AI, combined with fluency in their domain expertise, will be – as the Chinese saying goes, 如虎添翼 – like a tiger that gets new wings.

For the individual, it will likely open up new career opportunities.

For the profession, it will be a way to remain relevant to clients and attractive to new talents.

For the AI ecosystem, it will mean more sophisticated users who will spur demand for solutions and tools, which in turn attracts higher quality providers; you can imagine them being the best partners for the AI companies’ forward-deployed engineers.

Our ability to grow these AI bilinguals can therefore make Singapore’s AI Hub a stronger value proposition to international investors and partners, helping to create more good jobs for our people.

To get started, we are working with professional bodies to figure out the “minimum vocabulary” their members need for AI fluency, and to design practice-oriented training.

For accountants, this may be using AI to automate data sourcing and compilation.

For lawyers, to help search for precedents, reason across them, and construct sharper legal arguments.

Increasingly, many of these use cases will involve going beyond learning to write a prompt, to building and interacting with an agent.

But because such AI skills have practical use, these professionals may be more motivated to start learning and keep getting better.

We believe this is a better approach than an overreliance on generic theory-based training, which may not produce practical learning or lasting results.

Taking a step back, why is this work important?

It stems from our fundamental commitment to the vision of AI for the Public Good, which must also mean it is good for our workforce.

Most of us aren’t going to be model builders – these are the specialists, our AI creators at the top of the AI pyramid. Our sights are therefore set on the broad base of AI users, as well as a core middle of AI practitioners.

Amongst these AI practitioners, besides data scientists and machine learning engineers, we believe AI bilinguals – fluent in both their domains and AI – will bring unique value to our AI ecosystem.

Besides the domains I described earlier – accountancy, legal – there is another domain that occupies a special position. This is the domain belonging to the people in this room tonight – the tech and software industry.

In many ways, the tech industry helped to shape Singapore’s economic transformation. Today, our digital economy contributes 18.6% to GDP, on par with key sectors like manufacturing and financial services. It has consistently outpaced average GDP growth, and expanded employment more than other sectors.

This is perhaps why SGTech has no shortage of people and organisations to honour.

The Workforce Transformation award recipients this year are showing how to use AI for impact.

For example, at Acronis, a cybersecurity company, AI agents have become “team mates” in daily development work, helping engineers generate four times more code than before.

At ST Engineering, AI has transformed documentation, testing, and design – allowing engineers to prototype with unprecedented speed.

Stories like these point the way forward for our tech workforce.

They give us hope that we have tech solution providers who are ready when called upon, once again, to support Singapore’s transformation – this time through AI.

Whether it is our National AI Missions, the Champions of AI or the National AI Impact Programmes, I have no doubt many of your clients and colleagues will be seeking you out, to be pathfinders or partners in charting a way forward.

But if we are to be perfectly honest, piecemeal efforts in pockets here and there will not be enough. We need a better plan for the tech industry’s own AI transformation – a comprehensive response that addresses the concerns of all tech workers and tech organisations.

Whether we have felt it or not, AI is fundamentally changing what it means to be a tech worker or software developer. It is reshaping roles, career paths, and day-to-day work, and these shifts are unfolding faster each day.

Three groups will be affected in distinct ways, and a one-size-fits-all approach will not address each of your concerns.

First, the seasoned professionals. You have deep experience of how systems are built, what goes wrong, and why. The question is how we draw on that knowledge, to guide AI to build systems the right way – and how we help you stay relevant and valued through that transition.

Second, the newer entrants, i.e. our fresh graduates. You bring something genuinely valuable to the table – you think about problems differently, because you are AI-native. But I do not think that means there is nothing left to learn. The question is how we help you build the right experiences and intuitions to grow into seasoned professionals, when the traditional path for doing so is itself changing.

Third, those who are still in school. We need to look carefully at what is being taught, and whether it is the right preparation for the world you are entering.

That’s why at the Committee of Supply debate this year, I announced that we will enhance the TechSkills Accelerator programme, or TeSA for short, to help tech workers move up the value chain – from writing code to orchestrating end-to-end systems powered by AI agents.

We worked closely with partners such as AI Singapore, AI Centres of Excellence, leading firms, and government agencies to co-develop a curriculum that empowers our tech workers in the age of AI and directly addresses what the industry needs today.

The curriculum will cover AI-assisted coding, agentic AI, full-stack AI application development, and responsible AI practices.

In an AI-enabled world, TeSA will help seasoned professionals harness your experience, fresh graduates channel your instincts, and students build the right foundations. Because AI fluency, like bilingual skills, is not a test you pass once and then forget about. It is something tech professionals must keep developing and honing, at every stage of their careers.

At the same time, we know that individual reskilling alone does not lead to meaningful transformation.

Entire business processes need to be re-examined.

GovTech itself recognises this. Some of you have read the essay by our Chief Technology Officer Chang Sau Sheong, published on Medium in February this year. It describes the profound challenges ahead of tech organisations – but also the opportunities to re-imagine what’s possible.

Sau Sheong further observes that AI makes proofs-of-concepts so cheap and fast that we are quickly becoming “a world not ready for that much software”. The related question, and elephant in the room, is therefore “does the world already have too many software engineers?”

That reminds me of the idea starting in the 1970s that access to computing will bring about a “paperless office”. Remember that? What actually happened?

In fact, demand for paper surged, because early digital tools made creating and printing documents easier. It was only decades later, in the 2010s, that demand tapered off – when cloud systems, digital signatures, and better interfaces matured.

By then, new digital demands had emerged. Some organisations in the paper supply ecosystem successfully pivoted to serve the needs for data storage, collaboration tools, and compliance systems.

The broader lesson is that innovation rarely reduces demand immediately, often expanding total activity before substituting older forms over time.

But it does not mean we can afford to wait, and cruise along with the temporary surges in demand. Because the technology will inevitably mature, and some organisations will find themselves blindsided, unable to adapt in time.

Therefore, instead of being lulled into complacency, or fighting the inevitable that AI brings, would it be better for us to ride the wave early – and stay at its crest when demand shifts?

These are the kinds of structural questions we want to work through with industry.

That’s why, I have asked SMS Tan Kiat How, to lead consultations with the tech workforce over the course of this year – to listen, test our assumptions, and work through what a fuller response to AI should look like.

SGTech and other TACs have already been doing important work engaging with industry, and the Government will broaden these conversations to shape our policy interventions.

Our aim is to complete this work by the end of the year, so that we can respond with concrete measures as soon as practicable.

We are embarking on this exercise because we want tech professionals in Singapore to continue to thrive in this age of AI.

When we think of "Champions of AI", it need not refer only to companies and organisations. What if the entire tech profession is a “Champion of AI” – comprising of first-movers, early adopters, and the best-of-the-best among AI bilinguals who are genuinely skilled and confident in putting AI to meaningful use?

Is that a worthy goal? Why not?

Colleagues and friends, as you ponder these questions, take heart in the fact that Singapore has never shied away from change. Instead, we have always believed it can be and should be navigated with care – for the many, not just for those already well-placed to benefit.

The Government will play its part. But we will also need partnership, candour, and ideas as we work through what comes next.

Tonight, as we celebrate what we have built, let us commit to jointly shape an even brighter future for technology in Singapore. I look forward to being part of this journey.

Thank you!