Transcript of Fireside Chat with President Tharman Shanmugaratnam at Singapore Fintech Festival

Question: We know that technology is reshaping the economy. We've seen things like AI, digital assets, really pushing the frontiers of the global digital economy. But these advances in technology have also brought together tremendous promises and some precarious pitfalls. And this discourse has become top of mind for many governments and regulators at this festival, who will be dissecting some of these issues further. Do you think it is time for us to do a fundamental rethink as to how we conduct financial services and the likely impact of these really fast evolving technologies such as generative AI and digital assets?

I think financial services will be more impacted or more quickly impacted than other sectors, but this is going to be economy-wide and society-wide. That chart which Kristalina (Georgieva) showed, on the speed with which new technologies have acquired 100 million users, it was quite instructive. ChatGPT got 100 million users within two months. 

We always knew – it started off with sci fi, and then it became very serious thinking, and now it's becoming reality - we always knew that machines, and in particular now large language model (LLM)-powered chatbots, are going to take on more human characteristics. And we wondered whether you will reach that point of singularity, where the machine becomes equivalent or smarter to the human brain. There's no intrinsic reason why it can't happen. I don't think this is going to happen suddenly, where we wake up one day and find singularity. It's going to happen progressively. And I think it'll happen for certain human tasks much faster than previous technologies had done.

There was a recent study in the US that found that for 80 per cent of jobs, about 10 per cent of the work can be quite easily replaced by AI – and AI, you have to remember, is just part of a continuum of the digital revolution. That's not very alarming. But for a quarter of jobs, one out of every four people, it will be about 50 per cent of their tasks that are quite easily replaceable by AI. And that again is not fundamentally alarming. It may not mean that they lose their jobs, but their jobs are going to change in a fundamental way.

If I have to guess, I would say AI is going to be faster than previous technologies in replacing human tasks and enabling humans at the same time. There will be not dramatic change within three or four years, but I think in 10 to 15 years, this technological revolution is going to have a profound effect on workforce.

I think the basic attitude we have to take is, how do we use technology to help humans flourish - that's what technology has always enabled, it’s the key driver for human flourishing. But we have very important issues to think through, the distributional consequences, and must address those distributional consequences.

The earlier waves of technology - automation, in factories, logistics and the like - essentially replaced what was called routine and repetitive work. What's different about LLMs and AI is that it takes over cognitive tasks. It takes over tasks that are done by people with better education and in fact better incomes.

There may be something equalising in that. It's finally reached home to the segment of the population that does the thinking work that technological disruption is disruptive. It's disruptive, and it's becoming real. I can’t help wondering, as an aside, if one of the reasons why almost every day you read an article about AI and its disruptive effects is because the people writing them are part of the cognitive class.

So the first point I'm making is that this new technology is going to uplift human potential in general. The second point is that it's going to have very different distributional consequences from then past. But there's a third point too, which is that it might upend the traditional hierarchy of jobs. What you regard as a better job – and a job that deserves better pay - might change in the years to come.

The way we value IQ over EQ might change. Because there's a whole range of jobs that are thought of as requiring IQ that can be taken over and, in fact, done more efficiently and better using LLMs. And we’ll want to give more value to the jobs that require EQ, the jobs that require teamwork, and the jobs that require collective imagination.

Question: Let's talk about how we can square the ability to balance between using AI for the betterment of society, and really addressing the inherent risk associated with this technology. Because earlier this month, as per the Bletchley Declaration, we had 28 governments willing to come together to work on AI safety research, touching on scientific and international collaboration. It is very timely because there are growing concerns about AI safety, and the risk associated with loss of jobs and privacy. So Mr. President, how do you think we can really find that right balance using technology to allow humans to flourish, but also address these risks?

Singapore was an active collaborator in the Bletchley conference and declaration.

I would say first, look out for the worst things that can happen and try to avoid them. Matters of life and death in healthcare, for instance.  AI is going be a huge enabler and it’s going to open up a whole new field of healthcare. But on matters of life and death, you do need human judgement, human decision-making. We will use AI as a whole new tool for both diagnosis and treatment, but we still need human judgement, and we’ll need regulation, within countries and internationally, to contain the role of AI in healthcare. Contain doesn't mean holding down - we’ve got to use a lot more AI in healthcare. But we've got to make sure that decisions are ethical, in the interest of each individual patient. That's one of the examples of what can go wrong if AI is unregulated.

There are concerns with regard to national security, and also global security or mutual security. I'm not an expert in the area. But AI has the potential of leading to unintended catastrophic outcomes, outcomes not in the interests of the contending parties. And that too requires some rules of the game.

In financial services, I think if you're talking about a person's life savings, particularly an ordinary person's life savings, we got to be careful. AI again has great potential for improving financial advice.  It can provide a lot better informed and more customised advice. But it’s someone’s life savings, and an ordinary person's life savings where they don't have too much, you still need the human being and you need ethical rules of the game within finance to ensure that the right advice is being provided.

Question: What role then do you think multilateral development banks, the public-private sectors, investors and philanthropists can play to make sure that there's enough governance surrounding the use of AI?

It's early days. We're very early in the game of even thinking of how we can regulate this. And my cast of mind is not to be too all-encompassing, not to be too comprehensive. You can do a lot of thinking about it, but focus on the tasks that involve avoiding the worst. I would pay a lot more attention on that.

If you ask me about the range of risks we face – in health and financial savings, in security, I would say each of them requires focused attention on the worst things that can happen. It must involve an international coalition of countries. It's not just for the most advanced countries to decide, it must involve a broad range of countries.

And we have to think about a challenge that we are all going to face in the years to come, which is about politics, and how AI can be potentially very damaging to democracies in the years ahead. The potential for use of deep fakes, the potential for very sophisticated misinformation, is great. It's already being used. And when you combine that with the algorithms of the large social media platforms, which systematically lead to polarisation, it’s not as if we have a marketplace of ideas that leads naturally to people becoming more open-minded as you listen to different ideas. In practice, what’s been happening too much of the time is that people reinforce their views and their emotions. That requires regulation as well. The combination of AI and deep fakes with the social media algorithms is a very serious challenge to the flourishing of democracy - to maintaining open-mindedness and to ensuring that democracy is about the coming together of minds and hearts, rather than the coming apart.

Question: I want to talk about technological innovation for a pocket of society that I know you're pretty passionate about, that is the ageing population, and how technological innovation can assist the ageing population to thrive. This is a problem that's confronting many developed nations today - slower economic productivity and growth. You’ve said before, and you're the patron for the Digital for Life movement, you said that societies with ageing populations need not become pessimistic societies, or societies that feel that the future has limits. So how do you see technology and innovation allowing anyone including the seniors and ageing population to continue contributing to economic growth, while at the same time stay resilient amidst the changing demographic realities?

I do think that technology, including this latest wave of technologies we're seeing, is going to be a huge enabler people to live life fully in their later years. The opportunities to contribute throughout life are now vastly expanded.

First, going back to health care, because that's central in the minds of most people as they get older – the power and quality of healthcare is going to go up. And AI is one of the reasons why, in both diagnosis and treatment. I don't believe we're going to get rid of doctors, by the way - we might not need as many doctors as societies will otherwise need as they age, because AI can speed up diagnosis, but the role of doctors is going to remain critical because of the human decision-making that I spoke about earlier.

But what's interesting is what this means for the person who spends most of the time with a patient - the nurses and the other care professionals and assistants, in a hospital and outside in the community. Their jobs are about EQ as well as tasks that can be greatly enabled by technology. So when you talk about that upending that I spoke about, I think the role of the nurse and people involved in the care professions is going to be enhanced. They are going to have more tools. More information given to them to be able to provide that triaging, having that relationship with the patient that's increasingly going to be customised, in other words, for that patient individually. And they’re going to complement doctors in a new relationship.

Another opportunity is about continuing to learn. The beauty of today's technologies is that they help you keep learning through life. We can all do it. The tools will be in people’s homes, but it should also be about groups of people getting together to learn. Doesn't need to be classroom style, but groups of four and five people getting together in at places they enjoy, in a cafe or in community clubs.

It's changing basic conceptions - that you learn when you're young, you then work, and you finally retire and wonder what to do with your time. We can keep learning through life and the technologies now make it possible for us not just to learn but to stay active and to keep contributing.

I don't think we're already doing as good a job at this in Singapore as we should. We have to become a less ageist society. We’ve got to make the most of an older population, to enable them to keep contributing, and to retain dignity at work for as long as possible.

I was just reading something about Japan: a substantial proportion of Japanese firms, more than most other developed societies, employ persons above the age of 70. And when you visit Japan, you can see them up and about, and performing their tasks with some dignity. That's what all of us should be like, our society should be like.

And you have to remember that when we talk about EQ, it’s not just something you're born with. It's something that accumulates over time and the more experiences you have in life, the more of what we call EQ you have - it's just the range of experiences in human contact, and understanding the human condition, that comes with age. Our seniors are a valuable part of that human-sensing workforce.

Question: The Singapore Digital Society Report was just out, and it showed that 78 per cent of our seniors here in 2022 are now comfortable using e-payments as compared to 2020 when it was just 33 per cent. On the topic of learning, I want to come to another segment of the population. I want to get your solution to this perennial problem of the global tech talent crunch. This is a challenge facing many companies and many governments as well. We know that businesses need help to constantly innovate and stay competitive in the global marketplace, and for governments, they need highly skilled tech workers to develop and implement essential public services and address the challenges of climate change and cybersecurity. So what advice do you have for us to cope with the advent of new technologies? What can workers do to keep upgrading or refreshing their knowledge and skills?

Some jobs are going to be much more affected than others. In some areas, what is a shortage today might become a surplus.

We’re short of people with the skills required for cybersecurity, we're short on many AI-related skills. But what AI is also going to do is to replace a whole set of programming jobs, a whole chunk of the business process outsourcing (BPO) and customer services work. And this is a large number of people globally. It's good for countries like Singapore because we are short of people, we don't have many of our programmers.  

More worrying is that it will make it more difficult for developing countries to catch up through the BPO business, programming jobs, and exporting ICT services. A large swathe of those tasks are going to be taken over by AI, and it's already happening.

So we are short of some of the tasks that require the deepest knowledge and skills. But we used to be short of programmers, and I think we'd be much less short of them in future. But very importantly, we’ve got to ensure that anyone who is displaced can be reskilled, and use technology to enable everyone in the workforce to be more productive and to enjoy their jobs. Those are challenges that we have to address.

I think that the societies that have public-private-labour union coordination - German or Scandinavian or Singapore style - are going to be well-placed to address these challenges. You can't just leave it to individual firms and individual workers. It does require collective organisation, to provide an alternative for that large band of workers who are doing cognitive work that might be replaced by AI, in virtually every sector. We’ve got to provide them alternatives – either doing the same job differently and better enabled by new tools, or moving on to other jobs. It's a collective task and we need to think of ourselves as a social organisation to address this.

Question: It’s also important to have diversity and inclusion in the workforce, where anyone and everyone feels welcomed and supported if they want to pursue a career in FinTech or technology. What do you have to say about that?

I agree entirely. It comes back also to the point about the intrinsic. We need systems, we need to organise ourselves, that's the extrinsic part of it. But we need a culture where everyone has that intrinsic ability and desire to keep learning, and it doesn't matter what the education levels were when we were young. Everyone can keep learning, and the rest of society has to place value on the fact that each individual is making the effort to learn and is capable of doing something. And when you recognise that intrinsic quality in people, and you make it possible for them to keep learning through life, it also changes the way in which we value different individuals in our society.  

Question: This comes back to your point about how EQ is still going to drive that sensing quality that we need to find amongst citizens of any society. If we were to look at the Singapore FinTech festival, what role do you think the festival can continue to play in the international community in addressing or navigating some of the more complex and sustained global challenges?

This is something that sprung onto the scene eight years ago. We are one of three or four major gatherings internationally; there is the Point Zero Forum in Europe, there's a FinTech forum in Africa which we in fact collaborated with to establish, there are a few major gatherings internationally. No one place should do this on its own, there have to be a few nodes internationally. Singapore’s FinTech Festival, I would say, has got a very good composition of persons from around the developing world, as well as the most advanced countries. It’s distinctive in that sense. It's got very good Asian representation, but it's got people from every corner of the world, and I like that.

Diversity is a strength in FinTech and in innovation. And (Singapore) FinTech festival has to thrive on diversity.

Question: Given that we will have about 62,000 people come through these halls over the next three days, what are the top two things that you hope visitors to SFF will take away with them?

We should look for things that surprise us, which are the technological advances. But we should all keep in our minds the idea that we must be out to do good. To do good for either those who are still not included in financial services which Kristalina spoke about, or do good for people who just start off with less than life and need an uplift. So just keep in mind that this is all about doing good for society or some group in society.

And by the way - when I sat down I realised that this can of drink was called Ice Mountain - we should all remember the ice is actually melting on the mountains. And it's going to be the biggest challenge we face. We're reaching tipping points in the Earth system that are very dangerous, and we've got to use everything in human ingenuity, technology, and global coordination to avoid crossing them.