Transcript of Minister for Foreign Affairs Dr Vivian Balakrishnan’s Participation at the National Security Commission on Artificial Intelligence’s (NSCAI) Global Emerging Technology Summit via Zoom on 13 July 2021 at 2215hrs

National Security Commission on Artificial Intelligence Commissioner Gilman Louie: Well, it is great to be here and I want to introduce our special guest the Foreign Minister from Singapore. Vivian is a leader and a thought leader in all things digital. I think his experience, not only because of his representation in his role as Foreign Minister, but his prior role as Minister-in-charge of the Smart Nation (Initiative) is something we all can learn from. So let us have a round of applause to welcome the Foreign Minister and I will turn it over to the Foreign Minister for some opening comments before we have a short Q&A.

[Applause]

Minister: Thank you Gilman. I am the Foreign Minister of a tiny city-state in the heart of Southeast Asia. In order to be useful, I am going to take some liberties, make some provocative comments and invite some tough questions after this. First point, there is no doubt that Artificial Intelligence (AI) – the advent of computer systems that can see, hear, translate, speak, recognise patterns and play games – there is no doubt this is transformative both at a human level and indeed at a planetary level. The second point is that this tsunami is unstoppable. It is unstoppable because of the explosive increase in computing capacity, in global connectivity, in the explosion of data sets – especially with the Internet of Things – and of course the increasingly sophisticated mathematical models and the algorithms available. So it is not going to stop. The question is – which country or which regions are going to be at the forefront of it? What are the implications, both strategic and economic, of that contest? Second set of issues – what are the downsides of this explosion in AI? And third, what do we need to do about it at a domestic level and at an international level?

So let me start by saying I am going to duck the first question of who is going to be ahead. We can indulge in crystal (ball) gazing during the question and answer session. But let me quickly go through the potential downsides of this revolution. The first thing is to understand that AI is very different from the Industrial Revolution or indeed, even the agricultural revolution before. Let me give you an example. If you have a hammer in your hand, it is an extension of my muscle power. If you have an Excel spreadsheet, you basically extend the computing capacity of your mind onto a computer system, but it is still imperatively obeying your instructions. With AI, the difference now is that you have outsourced the means of pattern recognition, and even of the analysis and response to the system. Which means you are no longer programming the system; you are a coach to the system. And if you take the sporting analogy of a coach, the whole purpose of coaching is so that your student exceeds your capacity. There are a few critical implications of this. Number one, is that once you allow an autonomous system to define its own means and optimise its own methods, actually, values, ethics, morality, mission and vision on a human scale do not necessarily exist. It is worth remembering this – I know there has been a lot of talk about liberty and privacy – but bear in mind that AI on its own does not and is not restricted to human understanding of these concepts.

The second point worth emphasising is that in the early phase, I think we are going to see an erosion of trust. You see that even now in the very early stages; the lack of trust in authority and experts, the depletion of social capital even in the midst of a pandemic, the fragmentation of our political and social constructs into echo chambers with no real middle ground and an exchange of views. The third implication of this AI revolution is actually on the strategic scale. In a nuclear age, we sort of kept the peace through the concept of Mutually Assured Destruction. But when we move into an age where it will be autonomous swarms of submarines and planes and missiles, and in order to make those systems effective, ultimately you will take the human out of the equation, and there will be killing machines. At that point, what does the concept of ‘human deterrence’ mean? So I would posit that in fact, we are moving into a very unstable, potentially dangerous age ahead of us at least in the short term.

Now, understanding these risks, what are we going to do about them? Let me share a story. About six years ago, I had a side conversation with President (Barack) Obama. We were waiting for a meeting to begin and I asked him Mr President, what are you most worried about when it comes to cybersecurity? He looked at me and he told me – well, in state espionage we all do it, but we just object to commercial espionage. I thought about that remark and, frankly, I think there are some problems. The first point is that it is a remark that can only be made by a nation that is completely self-assured about its unprecedented lead in cyber instruments. And therefore, you can wave (it) away and say well state espionage is fine, it is just commercial espionage that I am concerned with.

The second problem with that statement, is that it also assumes that there is an automatic wall between the state on one hand and commercial companies on the other hand. If you think about it, both the state and the internet majors are all trying to erect walls. That is why you get questions about data flows and data localisation. At the same time, the truly multinational data pools lie with the internet majors, but they also are trying to construct walled gardens so that they keep the value of the data to themselves. The missing person, the loser in all this configuration between state and internet majors, is the individual who owns the data, who derives value from the data, and the economic implications of that. And now, somewhat belatedly, the United States – and here I will apologise in advance – but the United States is now suddenly questioning whether it needs industrial policy. The concept of industrial policy brings to mind the old Marxist concept of the commanding heights of the economy, that there are certain critical components of the economy that are of such strategic and competitive advantage that governments need to pay inordinate attention to subsidise, support and expand deliberately as a matter of public policy.

Again, we do not have time to go into that debate, but it is worth me reminding everyone, that the outcome of this contest will depend on which governments are able to provide enough attention and subsidies to education, to infrastructure, to enlightened immigration policy, to create a level playing field for fair competition between all commercial players, whilst protecting the interests of individuals and citizens. And getting the balance right so that you socialise the gains, and not just simply privatise the gains and socialise the downsides. And that, really, I think will determine the outcome of this contest.

For the rest of us in Southeast Asia, we have made tremendous progress in the last five decades because there was in effect a single industrial application stack for science, for engineering, for technology, for medicine, and that gave us huge advantages. If we now go into a world where at the advent of a new technology that completely rewrites the rules, we witness a fractured world without common rules, without a common stack of norms, values and rules of engagement, then we will run into a world which will be more tense and less prosperous. The speed of progress will be interrupted by conflict rather than competition. And so, speaking from Southeast Asia, we watch all this with great interest, grave concern even, and with anxiety. So what happens in the next couple of years is going to be crucial.

And on our part in Southeast Asia, we will remain open for business and we will remain open for technology diffusion. We believe that AI is another example of a global commons. For finance, we have the Bretton Woods Institutions. At a strategic scale, we have the United Nations. For climate change, we have the United Nations Framework Convention for Climate Change (UNFCCC). For oceans, we have the United Nations Convention on the Law of the Sea (UNCLOS). I believe we also need a new United Nations convention for AI; to make sure we develop it, we invest in it, we give opportunities for our people, and we pursue it in an ethical way. Hopefully, we usher in a golden age of peace and prosperity, rather than repeat the mistakes of the last century – where we went through enormous pains and troubles before we finally did the right thing. So let me stop there and take questions from you, and please be robust in your questioning.

Louie: First of all, thank you for those really important comments. I wanted to pick up where you left off on the importance of international rules and international standards. Singapore has always been a great leader in being in the right place at the right time. First, in its actual physical location being a leader in shipping and other forms of physical transits. And in your (previous) role as the Minister in charge of smart cities and digital initiatives, Singapore took the lead as being the nexus of how to deal with things like international cross-border transactions. That required a huge amount of cooperation, sometimes by countries that do not play well together. As you look forward in AI and the challenges that you outlined, any advice you can give other nations on how to move forward in an organised way to resolve some of these challenges? To establish what are fair and ethical standards of competition, and ways to resolve differences, particularly in these areas where there are social or economic differences on how you would apply AI?

Minister: Thank you Gilman. I think the first point is that all politics is local. All governments need to be able to explain to our own citizens that look, there is a revolution going on, but we are going to transform the education system. We are going to restructure the economy. We are going to look at safety nets, because there is going to be such profound transformation over the next few years. You first need to get domestic support for this transformation which is going to occur. So in a sense, US President Joe Biden's focus on building back better, on raising the level of confidence and hopefully consensus and congruence domestically, is absolutely essential. Now, assuming you have got political consensus and the seatbelts on for the roller coaster ride which you are going to get on, you then need to invest in the infrastructure which is needed. Everyone needs fibre to the home. There needs to be 5G for everyone. Internet access has to become almost a common utility, like water and electricity.

Next, you need to have an education system that works, not just from kindergarten to the age of 20. All of us are going to have two or three careers in a lifetime, and you have got to make sure those subsidies flow to enable us to retool and reskill for the next set of jobs that are emerging. You got to also have a system which allows you to build up your local workforce and be able to complement it with talent from around the world. If you look at Silicon Valley – if there was no enlightened immigration into California in the last 30 years, you would not have had the digital revolution that we have all enjoyed so far. On a global scale, as I said, we do need some international conventions. We do need to be able to interoperate, and we do need global finance to flow. Data is not just the new oil. Data is the new currency of the new economy, and it has to flow. But it has to flow in a way in which security is still protected, privacy is respected, and the value that is derived from it is distributed fairly. So there is a lot of pretty fundamental work that needs to be done. My point is that this should be done cooperatively, multilaterally, a new Bretton Woods in a sense, for the digital realm, the economy and for AI.

Louie: As you point out the importance of using these emerging technologies to better the human condition, Singapore has led the way in integrating these emerging technologies in everyday Singaporean life from healthcare to education to digital transactions. How was Singapore able to do all of that while still maintaining, you know, security and privacy and openness, and do so in a very reliable way? Can you share some of the lessons you learned, and what Singapore has learned in that transformation to the rest of the world?

Minister: Well, first thing I will say is Singapore is a very small place so in a sense it is easier for us. For instance, to lay a fibre or in fact two fibres to every single home and office in Singapore is relatively easier than if you were in a continental-sized country. But beyond hardware, it is about making sure that it is inclusive; that there is no neighbourhood that is going to be deprived of internet access; that all schools will have cutting edge technology; that all tools will be democratised and commoditised. And even as we digitise a whole suite of government services, that we make sure no one is left behind. And if you need to give people computers and pay for their broadband or give them smartphones, do so, so that we can move faster without leaving anyone behind. So again, I just want to emphasise that need for a domestic focus and to carry everyone along with you. And then after that, you have got to make sure you maintain trust. Trust is the coin of the realm. Without that, so much of all these fanciful plans that we can all dream of will not work. Once you can get to that level, then finance, services, and using this as an avenue to expand opportunities for everyone regardless of what field they are in – everything becomes possible. So we (Singapore) are fortunate to be a small place, well-integrated, well-connected, with top-level infrastructure and a digitally literate population that is willing to use those tools and get ahead. We are also lucky because we are in a unique place, both geographically and strategically. America has more invested in Southeast Asia than America has invested in China, Japan and Republic of Korea combined. Most people are not aware that America has so much invested here. At the same time, China is our biggest trading partner, and we are significant investors in the Chinese economy as well. For us, a world in which these competing giants hopefully get along well, is one with a lot of opportunities for us and our people. But we will have to watch this space with anxiety over the next few years. We do not control the agenda in Beijing and Washington, but we will try to be useful. We will be honest brokers. We will call a spade a spade. We will call it as it is, and in that sense try to be helpful.

Louie: Foreign Minister, thank you for your time and for your comments. Singapore is truly a world leader in many of these areas and we look forward to the continuing leadership you have in AI. Thank you very much.

Minister: Thank you.

.    .    .    .    .