Top
Please upgrade to a browser that supports HTML5 video or install Flash.Video-Preview-Image-r0-033.webp

The 2023-24 Arnold C. Harberger Distinguished Lecture on Economic Development

Sharing Tools

Link copied!

ABOUT THE TALK

"Framing the Effects of Artificial Intelligence on Global Economic Growth & Development"

The advent of Artificial Intelligence (AI), and indeed generative AI, demands a rethink of our approach to driving economic growth and development. At a minimum, the era of AI will likely disrupt the conventional growth models that have governed economic modeling and understanding for numerous centuries. In her talk, Baroness Dambisa Moyo will examine how the era of AI will have an array of effects, including:

  • On business: with implications beyond merely productivity gains and cost reductions
  • On society: particularly, if the benefits of AI largely accrue to holders of capital, rather than labor, risking widening inequality
  • On government: specifically, the changing role of the state, as the global economy confronts an era of labor-less growth, as AI technologies displace human workers and create a jobless underclass

 

ABOUT THE SPEAKER

Baroness Dambisa Moyo of Knightsbridge is a member of the United Kingdom’s House of Lords. She is a principal of Versaca Investments - a family office focused on growth investing globally. She also serves on a number of global corporate boards including Chevron, Conde Nast, and the Oxford University Endowment investment committee.

Her areas of interest are capital allocation, risk, and ESG matters. Dambisa is the author of five bestselling books on the global macroeconomy and geopolitics. She holds a PhD in economics from Oxford University and a Master’s degree from Harvard University.

 

ABOUT THE MODERATOR

Kal Raustiala holds the Promise Institute Chair in Comparative and International Law at UCLA Law School and is a Professor at the UCLA International Institute, where he teaches in the Program on Global Studies. Since 2007 he has served as Director of the UCLA Burkle Center for International Relations. From 2012-2015 he was UCLA’s Associate Vice Provost for International Studies and Faculty Director of the International Education Office. Professor Raustiala's research focuses on international law, international relations, and intellectual property. In late 2022, he published a biography of the late UN diplomat, civil rights figure, and UCLA alum Ralph Bunche for Oxford University Press.

 

ABOUT THE ARNOLD C. HARBERGER DISTINGUISHED LECTURE SERIES

Established in 1997, the Arnold C. Harberger Distinguished Lecture on Economic Development celebrates Arnold Harberger as an eminent scholar and teacher. The lectures provide a special forum for outstanding students of international economics and policy to present their views and research to the UCLA community and the public. 


 


Please upgrade to a browser that supports HTML5 audio or install Flash.

Audio MP3 Download Podcast

2023-24-Harberger-Distinguished-Lecture_otter_ai-audio-bo-cqa.mp3


Transcript:

Kal Raustiala 0:04

Good evening, everyone. I'm Kal Raustiala, director of the UCLA Burkle Center for International Relations, and it's my pleasure to welcome you to this year's Arnold Harberger Distinguished Lecture. And I'm very, very happy to in a moment introduce our speaker Baroness Dambisa Moyo. Before we get started, I just want to stress that we will run for one hour, and we will allocate time for questions from the audience, which you can submit through the q&a feature. So we'll we'll prompt you at some point during our conversation. But I want to urge you to put your questions in early we tend to get a lot of those, and I'll try to get to as many of those as we can. Dambisa Moyo will begin with some remarks. She and I will then have a brief conversation and then open it up to all of you. So that's the basic program. So with that in mind, I'm very, very pleased on behalf of the Burkle Center and our our co sponsor, the Center for Global Management at the Anderson school here at UCLA to introduce Baroness Dambisa Moyo of Knightsbridge, she is a member of the UK's House of Lords. She's a Principal at Versaca investments. She serves on many corporate boards, including Chevron Conde Nast, the Oxford University Endowment Investment Committee, she's worked and written on many topics, including capital allocation, risk, ESG matters, and others. She has written over five best selling books on the global macroeconomy and geopolitics. And she holds a PhD in Economics from Oxford, and a master's from Harvard. So Dambisa, it is my pleasure to bring you on and I look forward to hearing your remarks. Thank you so much for joining us.

Baroness Dambisa Moyo 1:53

Thank you. I'm delighted to be here. Thank you so much, Professor, Raustiala for such a generous and kind introduction. It is an absolute privilege and a distinct honor for me to join you on this occasion. And I'm deeply honored to follow many Nobel laureates, US Treasury Secretaries and eminent economists in contributing to this important lecture series that recognizes Arnold Harberger as an outstanding scholar and teacher, and celebrates his seminal work on taxation and welfare. I am grateful for the opportunity to share my thoughts on global economic growth and development at this time, and specifically, to offer a perspective that I hope will motivate sharper theoretical thinking and empirical work on the effects of artificial intelligence on the prospects for long term economic growth and development. As you just mentioned, my work sits at the intersection of three broad areas, public policy, as a member of the United Kingdom's House of Lords, business and private sector as a member of large global complex corporate boards and organizations. And in the not for profit sector as a member of the Investment Committee of the Oxford University Endowment. Well, these roles are distinct. Together, they helped me form a more comprehensive view of how AI, including generative AI could affect global economics and growth. It seems to me, this critical question warrants further serious investigation, especially given the growing consensus among leaders in the field of technology, that AI is a profound technological change, akin to the creation of electricity, the microprocessor, the Internet revolution, and even the discovery of fire. No doubt AI will materially change how we access public goods, such as health care and education. It will transform how we work, travel and communicate. Thus, artificial intelligence has the potential to revolutionize humanity and to reshape human progress. For example, in economic impact terms, the World Bank estimates that the digital economy already contributes more than 15% of world GDP. At a simple level, the debate on the effects of AI is characterized as a tug of war between two broad camps. At one end are those who are optimistic about AI and stress enormous productivity gains that could boost economic growth. For example, a 2023 report by investment firm Goldman Sachs estimates that generative AI could lift productivity growth by 1.5 percentage points over a 10 year period and drive a 7%, or almost $7 trillion increase in global GDP. Meanwhile, in July 2023 report by McKinsey, the global consulting firm, estimates that AI could increase corporate profits by $4.4 trillion per year. At the other end are more pessimistic views that AI could harm society by inducing mass unemployment, elevating cyber risks posed by errant state actors and unleashing machines capable capable of dominating or even destroying humanity. As academicians, economists, investors and policymakers, we need to find a balance to achieve the prospective risks and opportunities of AI. I believe that this new wave of artificial intelligence could impact global economic growth and development in at least four material ways. First, is on the economy. AI would likely cause a recalibration of the contributions to growth from the three key areas stated in classical models of economics. Canonical economic models show that growth is a function of capital, labor and productivity. Traditionally, empiricists note that productivity explains up to 60% of why one country grows and another one does not. Some early predictions project that AI could increase how much productivity now contributes to growth. The United States Congressional Budget Office currently projects a 1.5% productivity growth, which according to the Brookings Institute could compound to 33% productivity growth over the next 20 years. This means we must be open to the idea that productivity's contribution to growth could be far higher than 60%. Meaning Meanwhile, a report again by Goldman Sachs says that artificial intelligence could replace the equivalent of 300 million full time jobs, suggesting that our core economic growth model and understanding could change as productivity ends up contributing more to growth and labor contributes less. In the extreme, we can envisage a model of laborless growth, where the contribution of labor to growth becomes de minimis. This surely has implications for society and the role of government, which I will address later. Suffice it to say, laborless growth would mark a break from today's traditional economic thinking and modeling. Second, on business AI promises to be a catalyst for change beyond productivity gains and cost reductions. In particular, AI could materially upend business models, specifically altering how companies operate, recruit, and finance upon themselves. Take for example, the conventional venture capital business model. A VC firm raises funds and invest in startups in the startup ecosystem and early stage companies. It earns part of its revenue by charging a fee based on the total assets that it manages. These startups deploy the invested capital in their operations often to hire more employees. If AI reduces the need for labor, so that startups require fewer human workers to perform the same tasks, then employee costs will decline. This will in turn meaningfully reduce the cost income ratio, a key metric for many businesses. Therefore, early stage businesses will need less capital, venture capital revenues would fall, thereby disrupting their business model. The falling demand for labor will lead to less need for capital and this is a trend that would likely reverberate across other business sectors. Third, is the effect of AI on society. Thus far, I have explained that AI could have destructive effects on the labor force both through excuse me, both through the economy and through changes in business models. In essence, AI could accelerate us to a world in which more economic gains accrue to owners of capital than to the providers of labor, leading to worsening inequality. Markedly higher unemployment will further inequality access to education, health care and basic technologies such as Internet connectivity, and this could consequently worsen living standards.

Reduced access to key public goods can limit social mobility, undermine equitable growth, arrest development and even lead to civil unrest. Moreover, as well as worsening within country inequality, there is a risk of worsening between country inequality. In particular, gains from AI could accrue more to developed economies than poorer countries, thus widening the wealth gap between richer and poorer nations and reversing the convergence of recent decades. Fourth, AI could be the impetus for a fundamental shift in both the role and scope of government. Sure, this is an age old debate, the contest between those who believe in the economic growth and social benefits of big government versus small government, AI could tip us towards larger government towards more taxes more regulation, not least because of the declining workforce. Irrespective of left versus right political leanings, the prospect of a larger jobless underclass would create risks that demand a new approach to government. Furthermore, it will require refreshed thinking about economic growth and development. At a minimum, we should expect government to change how they approach two big policy areas: taxation and welfare. On taxation, we have established how an era of labor less growth would see economic gains accrue evermore to owners of capital and away from labor. This would warrant a shift to greater taxation on capital, and less so on labor. Specifically, this could mean a much higher corporation tax rate as companies generate excess profits from a smaller labor force. With regards to welfare, the threat of rising structural unemployment from AI, reinvigorates the debate on government support and specifically the need for a universal basic income. Essentially, a universal basic income would involve funneling the proceeds from higher taxes on corporations as transfers to citizens including payments to those who are permanently out of work. If, for example, unemployment were to rise to 25%, similar to the highest recorded rate in the US during the Great Depression in 1933. Government would have to channel funds to help households pay their way, excuse me through life and mitigate the risk of social instability born from rising poverty and inequality. A pessimistic view offered in 2013 by Oxford Martin School is that 47% of US jobs are at risk from technology. That is nearly half of all US jobs could be lost to automation. More recently, in May 2023, the World Economic Forum offered the more sanguine forecasts that global job losses from AI of 83 million could be offset by 69 million new jobs, leading to a net loss of 14 million jobs or 2% of the workforce. However, even this more optimistic view, would lead to pockets of very high structural unemployment, which could be skewed towards younger workers and poorer communities. Of course, these more calamitous trends or draconian responses may become moot, say if job diminution proves to be much less profound, and there is far less mass unemployment, or if productivity gains between capital and labor and across countries is much more widely spread, thereby limiting inequality and social disruption. Even so, the fact is, AI presents a challenge to how we model growth, the economy, business, society and the role of government. In effect, my remarks today are a clarion call for us as academicians, economists, policymakers and business leaders to think to rethink and possibly even rewrite economic models to ensure that they're equipped to work effectively in this new era. Thank you. I'd now delighted to be able to discuss this topic more thoroughly.

Kal Raustiala 14:55

Great, thank you so much for those remarks and for for raising so many provocative issues about a topic that's obviously on everyone's mind. And it's raising questions about governance, about economics, about livelihoods, and all of the above. Let me ask you first about this kind of arresting phrase, you used of laborless growth. Here in the United States, we're at a moment where organized labor has been more active than we've seen, in many years, many large strikes throughout the United States, even here in the University of California. And, at times, those labor actions have noted that the role of AI or the fear of AI, so I'm just curious whether, first, if you could expand more on what laborless growth would actually look like, in the coming years? And then do you think there's any connection between the renewed attention to unions and the renewed power of unions and the sort of specter of AI that seems to be looming ever larger?

Baroness Dambisa Moyo 16:00

Thank you very much for that question. I think, to put it in context, you know, whenever we've had enormous technological disruptions, I touched on a few of them. And of course, the Industrial Revolution being a classic one, we have seen shifts in the labor market. And so in that respect, any changes that we're seeing are not necessarily surprising. So if we think back, and just to give you a couple of historical data points, in 1900, in the United States, about 60% of the workforce was involved in agriculture. Today, that number is less than 3%. And we know what's happened as economists. And as historians, we know that a large swathes of the workforce moved into manufacturing, as agriculture became more automated, and then out of manufacturing into services. And now there's this question about what does it look like to move out of out of services into a more r&d focused type of of labor force, of course, there are real, real disruptions and changes that occur. And if you think about it today, we now have about 80% of the workforce, working in service sector, and as I said, less than 3% of people working in in agriculture, about 18% in manufacturing. So fundamentally, it's not surprising and ought not to be surprising to businesses, but also to governments that we will see this type of shift in the in a period of of technological advancements. I think it's also worth noting that a lot of the shifts that we're seeing in the labor markets are not purely only because of AI, you're right, that we have recently seen a lot of action by by labor unions. But I would hazard a guess that it's quite specific. We've seen it in the automobile industry, we've seen it in Hollywood. But we haven't really seen it in a more broad based sense. And I think that might reflect the fact that labor participation rates over the past 10 years. So this is people's willingness to work has actually been declining pretty quite precipitously. Over the last decade, we have seen a bit of a an improvement recently, but that has been on the decline. So all of this is to say that, you know, we could take the well known John Maynard Keynes' prediction from the 1930s, that by 2030, we would be we would have a 15 hour work week, you know, looking at the restructuring of of work through work from home. The labor participation numbers I mentioned which were already declining, before the pandemic hit in earnest. I think that there are, it is a confluence of factors that emerging but given the focus of my talk today, on AI, there's no doubt about that AI could be quite a disrupter. And as I mentioned, in my speech, there's some data already out there that that's occurring. Just to make the point that I did make in the speech, it doesn't necessarily mean that there's net zero or net negative, we could see new economies emerge. And the question is, can those be things be scalable to the point of driving economic growth? And I think it's a bit too early to tell at this stage.

Kal Raustiala 19:29

Great, that's really helpful. So you, you sort of raised the point about agriculture that, that I've, I've heard noted before about this sort of dramatic change. And obviously, if you go back even further in history, virtually everybody was involved in agriculture at one time, in humanity. That's all we really did. And now it's this tiny portion in the developed world. So, you know, that sort of raises this common trope in economic thinking about you know, this time is different, and the response to this time is different. We actually had Carmen Reinhart give this very lecture a few years ago, and that topic came up. So is this time really different? So I guess I'd be curious, just to hear more about, you know, this is this is sort of the most central question about AI that I seem to encounter, which is the, you know, the question of, is this really distinct from these other transitions that you mentioned in the past? Or is it just yes, it'll be a rocky, rocky transition from something today to something tomorrow, but at the end of the day, this is the way that economies have always worked. We've always had technological change, we've adapted. Certain people, certain sectors have been harmed. The manufacturers is kind of the classic example. But that's okay. They're gone. And we have something else in its place. So is this time different?

Baroness Dambisa Moyo 20:51

Well, I think your answer kind of hints at at the issue, which is that we don't know what the sector that will absorb large swathes of human workers might look like. Now, when I speak to my friends in Silicon Valley, they say just because we don't know what it is doesn't mean that it's not coming. But unfortunately, with my public policy hat on, and we're not in a position to regulate or to oversee public policy on a hope and a prayer. And we would need, we would like to see what that sector might look like that would absorb young people and people from from disadvantaged communities. You know, the fact that the ILO, the International Labor Organization estimates that we have about 100 million young people. So between the ages of 16 and 24, who are out of work is a problem, particularly given that many developed developing countries, where 90% of the world's population lives are skewed to young people. And so if we have to deliver a changing structure to changing opportunities to this young group of people, we need to be able to know what that sector is. And right now, although there are hints of what it might be, I mean, it could be anything from the care economy, we have people talk about social media influencing, etc, we need to be certain that that it's going to be scalable. So it is, you know, we've got over 8 billion people now on the planet. But we also need to be sure to manage the demands of skills around that. And so, you know, if you look at the OECD PISA report, which comes out every few years, which ranks students around the world, on their mathematics, science, reading, and the numbers are quite disheartening, particularly from developed countries, which we've seen slide down the ranks over many years. And so the ability to say, Oh, yes, the sector that's going to replace services is probably going to be r&d is wonderful in some respects, because that could not only speed up the delivery of public goods, like education and health care, but it also could reduce costs. And we've seen this happen, the cost curves for technological inputs, so things like food production, transportation, telecommunications, those costs have materially come down, which is good for consumers, it's good for society. But to see those things happen so rapidly, and not necessarily see what sector is going to absorb the workforce means that we are in this very delicate place, I might just add one other piece of information, which is that there is a concern among some people in Silicon Valley, which is, is basically popping up in the minds of a lot of public policymakers, which is this whole idea that the although we've seen enormous gains from productivity in, you know, in social networking, for example, we haven't yet seen those type of gains, really, in a number of key sectors, as I said, which is things like health care and education, nowhere near have we seen that sort of comport and a lot of the reason for that is, the argument goes, that is because of regulation. So regulation is so complicated, and, you know, often required, but it can be so difficult to import new ideas and new technologies. I mean, a very simple example of this would be that AI technologies could be quite disruptive, but quantum computing will be quite disruptive, around questions of privacy. And so there's that trade off between do we want to see these massive healthcare gains? Yes, we do. But are we willing to do it at the expense of protecting privacy, or that becomes a big issue and so because of the regulation around the health care sector as an example, we see a sort of drag, or a slowness by which public goods as again education, health care adopt some of the bigger promises of AI technology in general. And there boy, we lose out on those productivity gains. So I would just throw that out as well as being a constraint to some of the productivity gains that and thereby making it perhaps not very different from the past.

Kal Raustiala 25:19

Yeah, I mean, that's interesting. I mean, I think I work, for example, in a arena university teaching where there's virtually no productivity gains, I think, in hundreds of years. And I know, there's a classic example of, you know, the, the orchestra, the symphony, and you still need just as many musicians, you know, I can only teach X number of students, and it doesn't really get any better or faster doesn't matter. So there does seem like there are the sectors that are very resistant, and those show huge kind of cost increases over time. But I guess those are, those are generally the outliers, right. And so what seems to make AI different than some of the past technological changes, at least in the kind of the discourse around it is that it's striking at creative work in a way that's different. So while while I can't teach in a more productive way, perhaps I can write in a more productive way, or perhaps my writing will be totally eclipsed. That's what seems to make, you know, ChatGPT and these other models, so frightening to so many people. So you referenced earlier Hollywood and we were talking before kind of in the greenroom about the fact that there was just this huge strike of the writers, television and screenwriters. And one of those big concerns was the fact that AI could be used to write scripts. And this is just one of many examples where a traditionally high paying creative endeavor seems vulnerable in a way that's never happened before. So can you expand on that? Is that something that we should expect more than? You know, what, what are the implications of that, given that those are white collar professions that often require a lot of investment? To get into?

Baroness Dambisa Moyo 26:59

You're absolutely right. And, you know, without making any sort of broad claims on on people's on the sort of cost income ratio? You know, obviously, there are many industries, which people make place bets, they think that a movie is going to do well, and it doesn't. And so that's part and parcel of how, you know, I guess, and again, I'm not very familiar with the, with the film industry, but I would imagine the people who are investing capital to try and generate certain returns, and that's what they'll be looking at, is there a cheaper way to deliver it? I mean, this is a very public argument. But but maybe what I'd love to do is just remind us ourselves that, you know, we've had a good 10 years, probably 15 years of enormous technological gains, and many of those have fallen flat. I mean, if you think about the promise of NFT's or of crypto Internet of Things that we've had the sort of some would call fads, come and go. And with that was lots of concern about what the diminution of jobs and what that would mean for society. And I think it's worth just remembering that to the extent that we could see a material change, it would probably take some time. And that's what history has told us. We, you know, this is one of the puzzling questions that many economists are debating now is why is it that we've been living in this era of technological advancement, and, you know, unseen for many decades, and we haven't seen the productivity gains that we would have expected. You know, I will also point you to a period in during the, the industrial revolution, that's called the climacteric period, where, although there were enormous gains and enormous technological advances in railway, and you know, obviously, transportation, but also speed with which people could produce goods and services, materials, etc. There was a period of about 10 years where productivity just fell off. And to this day, economists and historians can't explain what happened, how is it possible that we have all this enormous improvements in technology and not and not see those gains? I'll just say one other thing, just to underscore the point of how long these adaptions adoptions and adaptions actually take place. You know, the classic one is is Benjamin Franklin. So from the moment his kite got struck by lightning, and he realizes we can create electricity. I mean, it took I would argue, you know, centuries to actually deliver on that promise of electricity to put it into homes make it usable. So these things take a long time and I you know, I It's this is probably shows my, my is my age, but since all the euphoria of ChatGPT earlier this year, you know, I'm not sure that a lot of people are actually how many people have actually embedded it into their work programs. And we're waiting to see what happens with Microsoft co pilot which launched on November 1. I think there's a lot of interest in that we're expecting by some numbers 25 to 40% Increases in productivity. But you know, who knows? And I think there's, it's a lot, a lot to ask for a whole system to change. And so to the extent that there are any changes, I would argue that it's not going to happen anytime soon.

Kal Raustiala 30:28

Great. Let me turn before we go to audience questions, just to kind of the global dimension. So you mentioned in your opening remarks that you might expect or we might expect to see the existing inequality in the world get even worse. And we've seen many examples around the world of some countries doing incredibly well over the last 50 years, I was just in South Korea. It's an example of a country that's really had kind of astonishing growth over the last 50-70 years or so, and then other countries that have really stagnated. And so I'm just curious if you could expand more, especially given your background working on issues of aid and development, especially in Africa. What do you see as the likely path for the global south in particular? Will those divides grow? You sort of said that, but just in what way might we expect to see that?

Baroness Dambisa Moyo 31:17

Well, again, you know, I would just caveat, the point, I mean, I'm doing very well as an economist saying on the one hand, and on the other hand, but this is really the murky area of AI. I mean, I think that there's an argument for every argument that says, Oh, my gosh, we're going to have AI, and the gains are going to accrue to those people who have investments in AI technologies. As we get into this world of a laborless growth model, well, that smacks of of harming emerging countries that don't have capital invested in in the way that you might see an open AI or Microsoft or Google etc, who have the capital to invest in these platforms. But on the other hand, we are living largely in a in a globalized world, I say largely, because as you know, since 2007, when globalization peaked, we have seen a diminution of globalization and trade, capital flows, immigration, and multilateralism. And so there is a question of, in particular, the splinter net, this whole idea that there'll be sort of rival camps of technological platforms that could actually net net, you know, go against this whole idea of having a flat world, where actually doesn't matter where the AI is produced, it's it will be it can be deployed anywhere. But you know, it's at the risk of sounding sort of, again, sort of living in a murky world, I think it's too early to say, but if I were working on on with emerging market countries, which I do, to some extent, I would be worried about the education piece. So what kind of skills are the next generation adopting? What are they learning, and I would also be very concerned about this labor versus capital tension, which would put the emerging markets at a disadvantage, by and large, I mean, of course, within emerging markets, you have countries like China, you mentioned South Korea, which are more advanced economies, which do have the capital to invest in this space, but they're, you know, the poor countries, in South America, Africa, parts of Asia, I think would be, could be put at a disadvantage. And we haven't even talked about the sort of rogue risks. But those are sort of well known and assumed.

Kal Raustiala 33:39

Say a little more about that rogue risks.

Speaker 1 33:42

That's cyber. The costs of of sort of state actors, you know, with different ideological beliefs, creating additional costs of doing business of trade of capital flows. And, you know, we've seen the United States government put out explicit rules that they don't want American firms to invest in Chinese technologies, we could see a lot more of that. And again, that would contribute to this sort of Splinter net or fractioning of the global systems precisely at the time where we would prefer, as many people have argued a system that's much more collegiate much more unified, in terms of mitigating for threat, threat actors, whether their state or rogue actors, and I think that's that could be quite problematic.

Kal Raustiala 34:35

Great. So let me go to audience questions. And there's quite a few interesting ones. So I'll start with this. So the question is, with this increased development and use of AI, do you see the potential for monopolization of technology and information within these industries? What is the impact of such monopolization, for example, through patenting or so forth? So well, you know, that is obviously a hot issue here in the United States. I can't speak for the UK. But concern, we're seeing that out of the, you know, out of the Biden administration, for sure renewed attention to the threat of monopoly. How does AI fit into that?

Baroness Dambisa Moyo 35:15

So it's a great question, mainly because even before AI, sort of reared its head recently, and again, we're using the term quite loosely here. I mean, there many people would say a calculator was already artificial intelligence. So let's, let's just put that in context. But even before we were in this new era of generative AI, and thinking about the possibilities, there were already real concerns about about sort of how we would think about the the nature of of AI, and what government could do to curb it. So for example, you're right that with in terms of monopolies, we were already starting to see many sectors trend in that way. And I won't give a full history lesson. But I will say that if you go back to the late 1800s, with the Sherman Act, which was supposed to bust these monopolies, we've actually come into the light, we came into the latter part of the 20th century into early part of the 21st century. And we found that in most sectors, in banking, pharmaceuticals, airlines, technology, we ended up with monopolies, or I should say, worst monopolies, but at best oligopolies. So each of those sectors had just a handful of dominating businesses. And so that's what the Biden administration is reacting to, in particular. But if you think about what AI could do, I think it could accelerate to a handful of sectors and actors in different sectors that could lead to a monopolistic state if we don't have a much more aggressive, regulatory environment. And, look, I personally am not really a big government type of person. But I do understand the need for regulation. But I think sometimes it can get a bit extreme, and we want to make sure that we are mitigating for risks. But we're also not doing it to the point of choking off investment, I often talk about the difference between the US and Europe, on things like energy and the climate crisis. A lot of the conversations that I ended up with in Europe tend to be much more about risk mitigation. We need to curb co2, we need to shut down demand etc. And true, there is a role for for risk mitigation. But we can, we can grow our way through through having so much risk, the same conversation in the US contexts has tended to be quite different, which is much more recognizing that we do too much more aggressively pursue investment. And think about the whole suite of new energies, renewable energies, energies that can address the climate crisis. And I think that tension between risk mitigation versus investments very important as we think of AI, because I think the the sort of fear of monopolies can become so extreme that we kill off innovation. And not only does that put us at a disadvantage within a country, but it also obviously, at a time of geopolitical fissures, it will put us at a distinct disadvantage, when we are competing militarily, technologically and economically with other countries that have may have a very different approach and view to some of these issues.

Kal Raustiala 38:49

And so broadly, you sort of prefer the American approach to the European approach that you just laid out.

Baroness Dambisa Moyo 38:54

I do. But that's my nature. And again, just to be clear, I'm not saying that there's no role for government, I'm not saying that risk mitigation doesn't matter. But we can get so focused on mitigating risks, not withstanding the size and scale and urgency of an issue that we actually create more damage. You know, COP 28, is in a couple of weeks. It will be very interesting to see how those talks go. Because I think fundamentally, the whole climate debate gotten to a place where it was very much about risk mitigation, scope, one, two, and three. And, again, not withstanding the urgency and the scale of the problem, I think you can lead into these corners where there's less sort of incentivization and encouragement of investment, and people are spending a disproportionate amount of time on on curbing risks. But in many developing countries in particular, people will not prefer that route, especially policymakers will not like that route because it's putting people's economic livelihoods in challenge so and in parallel. So I think this is true for AI also, that, you know, if we pursue a lens of oh, it has to be risk mitigated to the nth degree, because of what it might become. I think that there are a lot of benefits to cancer treatments and education, public other public goods or on technology, infrastructure, etc, that get lost, because we are looking at the proverbial glass half empty, as opposed to it being a tool for us to continue to improve human progress.

Kal Raustiala 40:30

Thanks. Thanks. So. So next question is to what extent would a declining global population or declining populations in certain countries and regions eg Japan, I'll add, Korea, I think has now the lowest birth rate in maybe in the world? How will those declining populations mitigate concerns associated with AI and laborless growth?

Baroness Dambisa Moyo 40:55

Well, I tend to think because I traveled to over 80 countries around the world and lived in many places, I tend to think of the world as a global system. And, you know, picking Japan I think is quite interesting. I've been to both Japan and South Korea, because they are quite closed societies, particularly Japan, and resistant to having immigration. And so that might limit their ability to sort of replicate their the progress that they've seen over over several decades. You know, because I see the world as being a sort of global system. I think that AI and the fact that world's population continues to grow, we're at 8 billion and according to the UN, will continue to grow at a rapid clip until 2100, there'll be about 11 billion people on the planet from 3 billion in 1960s. And so because that that curve continues to move up, even though in some places, it is materially slowing even in the emerging world, I do think that that net net, we're not in a bad place. vis-a-vis the population. Japan has is fascinating to me, because again, not withstanding the bad headlines that they've had for two decades about slowing growth and concerns about stagnation. We've actually seen their per capita income at worst, they're flat. But in some cases, some estimates say it's even ticked up a little bit. And if we're interested in really trying to improve living standards for your population than having a world where they've had slow growth, no doubt about it, and now weakened currency, etc. but still being able to have a per capita income at a certain level that is, you know, again, potentially growing, but certainly flat. That sounds like a great deal, especially on the ground in places like Japan. No, no surprises. You've seen a lot of investors flocking back into Japan, like Warren Buffett was very public about his investments. And I think it's because there is a sense that the Japanese government can thread the needle of continuing to improve people's living standards over time without necessarily having this enormous growth spurt.

Kal Raustiala 43:10

I mean, Japan does seem like a bit unusual. I mean, I think there's there's some economist joke about four kinds of countries develop developing Argentina,

Baroness Dambisa Moyo 43:18

and Japan. Exactly.

Kal Raustiala 43:20

So there was something strange about Japan. Okay, so next question is a bit of a throwback to your earlier work. So has your view on foreign aid changed at all since you wrote your book Dead Aid? And I suppose you could tack on does AI change any of your views about development aid?

Baroness Dambisa Moyo 43:39

So thanks for that question. No, my views have not fundamentally changed. And in fact, if anything, I would argue things have gotten worse, or perhaps more urgent to, to revisit new interventions to help support the poorer and poorer nations, just to be clear to those people who are not familiar with my previous work, which is now I believe, about 15 years old. And the argument I was making was that we should support poor countries, but that just handing out cash that goes from poor, poor people, essentially, through taxation in rich countries and ends up in in poor countries, is not the most efficient way to create economic growth in a sustained way. And I think part of the reason that my my, my concerns about these aid programs has not changed is because we are now in a world where the donors, the United States, Europe, in particular, are have a much worse fiscus. They are massive deficits, huge amounts of debt. That is, you know, obviously now. The higher interest rates they have to pay higher interest expense. I believe the United States now is spending more money on interest than it is on defense. With the cost of living crisis, a lot of challenges in terms of Medicare, Medicaid and other sort of Social Security, other types of entitlement programs being put at risk, lack of infrastructure. I mean, there's a whole list of challenges in places like the US and Europe. And that was precisely what I was cautioning against, that we will be putting developing countries or making them vulnerable to, to donor fiscus was not something that was sustainable nor was it desirable, particularly since we do know that countries that have performed like South Korea, which was poorer than many African countries in the 60s, but now, I mean, as you say, look at it now, have not relied on aid, as a key part of their growth agenda, of course, keep South Korea did was an aid graduate, and did have some aid, but it's about it being short, sharp interventions, not open ended in perpetuity. And so fundamentally, my idea on aid hasn't fundamentally changed. The question quickly on AI, I do think AI could be, could be a game changer. But again, if people are being skilled or trained up in what I call outdated, or declining sectors, and they wouldn't be able to compete in a world, that's going to require much more science, mathematics, technology, I would argue, in that sort of AI world, of course, there will be jobs that are not science based. But still, you need to have some technological ability, but also a reliability on technological platforms. You know, many countries, not just in Africa, but in the emerging markets don't have reliable energy. I'm speaking to you in California. So you understand the blackouts and the challenges around a consistent energy. And that would mean that even if you had a great business idea, and you had the skills to deliver on some AI platform, if you don't have the broader infrastructure, which is often happens in poorer countries, you wouldn't be able to see the scale of the gains that we would hope to see.

Kal Raustiala 47:13

So there's an additional question that's kind of on point you a little bit answered but just to expand a bit. So the question was about the impact of AI and populations in Africa living in extreme poverty already on the margins? It seems like one of the things that you've emphasized in just now, for example, is the need for education and educated workforce. So, you know, maybe if you want to say more about the particular implications for Africa, that's great. But I'm also curious, just what is the role for educators? Is education, broadly understood, going to be the most important thing that governments can invest in in the coming decades? That sort of seems implied by some of your comments today, but tell me if I'm getting that wrong.

Baroness Dambisa Moyo 47:56

Now, I think education for sure is critical, which is why, you know, I think we ought to be somewhat frustrated when we hear that there are few productivity gains being had in sectors like housing, or health care and education, because of sort of regulatory capture, that we get so protective of these sectors that we don't see those gains, because those are, those are places where we really ought to see a lot more a lot more gain. You know, you're right. I've kind of touched on the Africa point, I think that it worries me slightly as someone who's been on the planet now, going on 55 years that we're still talking about Africa as a sort of stepchild, I kind of hoped that the new generation, the language would change the sorts of innovations and hope around resolving some of these issues with change, it's quite disheartening to be at midlife and look back and feel like a lot of the language still is, is sort of kind of archaic and smacks of the sort of 1970s or even 1940s. in how we address the Africa problem. I you know, my my fundamental view is it's not some big mystery. We've got a young talented population. There's the issues around needing to deliver on infrastructure, reliable health care, reliable energy at a cost effective way that sustainable and reliable education. It's not some big Jack in the Box mystery. It's about poor execution, I would argue. And it's true, not just in far flung places like Africa, it's true within the United States, if you go to some of the pockets of you know, even in Los Angeles, but in other cities in the United States, where we have simply failed to, to invest and to execute on a bigger agenda. You know, your question about prioritization, whether education is the most important thing, it's such a great point, because if you had asked me this question 10 years ago, or even 15 years ago, I would have said, absolutely you start with education. But as somebody who spent more time now with public policymakers, I would say that your your question is really at the heart of public policy, which is essentially about trade offs. We can't be all things to all, to all policy requests and demands. You know, I don't want to belabor the point. But if I may, for just a minute, if you're given $100, and you're in a poor country, you could choose to put the full hundred on education, health care, infrastructure. But if you put the full hundred on education, then what if people are too sick, they can't come to school, there's no, they're healthy, but there's no road to get you to the classroom, well, then you don't solve that problem. You know, it's a little bit like Rock Paper, Scissors, because you know, you solve the healthcare problem, but then people aren't getting educated. So then public policymakers say, well, then maybe we should put a third, a third, a third. Well, that might be fine as a way of keeping things ticking over, but it's not going to set the world on fire. So you're not going to get those big boosts. Which is why again, something like technology, and AI could be a real game changer if we're able to execute if we're able to, to, to really scale some of the the education and healthcare gains, you know, that we've talked about already. So I think it's much more complicated to say, which comes first. You know, I this is the this is the rub. This is the centerpiece of what public policy does is trying to is taking finite resources, and trying to deploy them in a way that we can get the biggest boost. And that will forever be a challenge for government, as I'm intimated in my speech.

Kal Raustiala 51:46

Great, great. Okay, so next question, probably final question. AI is being developed and driven by leaders with similar backgrounds, visions, and blind spots, what do you think are the most promising guardrail tools at our disposal, whether in policy, regulation, business, etc, to ensure this technology is deployed to achieve a net positive?

Baroness Dambisa Moyo 52:04

So to me that it's ultimately about skills, you know, we are facing right now a world of a lot of challenges, whether it's its war, its disease, we have a cost of living crisis, we have global fissures, we have the highest number of refugees on record, according to the International Rescue Committee, we have about 80 million refugees and displaced people in the world. When when we are sitting, all of us sitting in different rooms, making choices and decisions about, you know, whether it's public policy, or business or NGOs, we need to understand that, that we have to make decisions that are. And by the way, when I say make decisions, I don't mean just on paper, I'm talking about actually executing on decisions based on the best ideas. And so, you know, there's I like to say to people, nobody is going to do you any favors. It's not a favor world. And essentially, we're trying to find solutions. And so the best solutions are going to win. So I always focus on, on making sure that people have a very, very solid understanding of issues, not just their understanding of what the issues are. But understanding counterpoints. You know, I know that's kind of not fashionable anymore. But, you know, if you don't understand people's counter views, you're not going to be able to compete, you're not going to be able to contribute in a very global world where you will encounter people who have different views, fundamentally ideological views. And so I think there's too much sense, perhaps that you know, that there may be people behind the dark curtain that are discriminating, and I'm sure I know that they are, those people do exist. But I wouldn't say that that is a of my priority. My primary focus when I'm dealing with energy issues, dealing with questions of poverty, dealing with questions of inequity, I'm trying to find solutions, I will take the best idea wherever it comes from. And so the more people can come up with brilliant ideas, things worth road testing to solve climate change, solve inequality, figure out how to create growth, I really don't, I'm not really hung up on where they come from. I'm much more interested in making sure that, that that, you know, we get encouraged people to have the skill sets to form those from those views. That's what innovation is. That's what got us to where we are today. You know, wasn't too long ago, there was outdoor plumbing in the United Kingdom 1970s. Things can change very dramatically. But we need to be solution led and not sort of problem led.

Kal Raustiala 54:47

Right. Well, thank you so much. And I really appreciate you zooming in from London. I know it's very late, and we're at the end of our hour. So again, we're honored to have you and we hope to have you back.

Baroness Dambisa Moyo 54:57

Thank you very much. Thank you.

Kal Raustiala 54:59

Yeah, take care. Everyone

Transcribed by https://otter.ai



Please upgrade to a browser that supports HTML5 audio or install Flash.

Audio MP3 Download Podcast

2023-24-Harberger-Distinguished-Lecture_otter_ai-audio-bo-cqa.mp3